Crystal Violet Staining for Colitis‐associated Neoplasia‐guided Delineation of Ulcerative Colitis‐associated Neoplasia for Endoscopic Submucosal Dissection: A Four‐case Report

ABSTRACT

Ulcerative colitis‐associated neoplasia (UCAN) is often difficult to delineate because chronic inflammatory changes obscure lesion borders. Accurate assessment of lesion extent is essential for curative endoscopic submucosal dissection (ESD), yet conventional white‐light imaging frequently fails to clearly define lateral tumor spread. We report four patients with UCAN in whom crystal violet staining for colitis‐associated neoplasia (CV‐SCAN) was used to delineate lesion extent before ESD. All patients had ulcerative colitis in clinical remission and underwent surveillance colonoscopy. In these four patients, UCAN was identified as CV–neoplasia (CV‐N) areas during CV‐SCAN, whereas lesion borders remained indistinct on conventional white‐light imaging. After histological confirmation by targeted biopsy, a repeat CV‐SCAN was performed to determine lesion extent before ESD. In three patients, the CV‐N‐positive areas corresponded well with the pathological extent of neoplasia, and en bloc curative resection was achieved. One patient had a positive horizontal margin after the initial ESD. During an additional procedure, residual UCAN was again visualized as a CV‐N‐positive area and was successfully removed by additional ESD. No procedure‐related adverse events or local recurrences were observed during follow‐up. These cases suggest that CV‐SCAN may facilitate delineation of UCAN before ESD and may help identify residual neoplasia that is difficult to recognize using white‐light imaging alone. CV‐SCAN should be considered an adjunctive technique that complements conventional endoscopic modalities rather than a standalone approach for lesion delineation.

1. Introduction

Ulcerative colitis‐associated neoplasia (UCAN), which includes both dysplasia and colitis‐associated colorectal carcinoma, remains a major complication of long‐standing ulcerative colitis [1, 2]. Endoscopic resection is recommended for visible dysplastic lesions when complete removal is feasible; however, accurate delineation of UCAN remains challenging because chronic inflammation, scarring, and mucosal distortion frequently obscure lesion borders [3].

Endoscopic submucosal dissection (ESD) enables organ‐preserving treatment of selected UCAN lesions but requires precise assessment of lesion extent before resection. Although image‐enhanced endoscopy, magnifying endoscopy, and chromoendoscopy have improved the detection and characterization of UCAN, delineation of lesion margins remains difficult in some cases. Therefore, additional techniques that complement conventional endoscopic evaluation may facilitate the determination of resection margins.

We recently developed CV‐SCAN (crystal violet staining for colitis‐associated neoplasia), a novel staining method that identifies UCAN as CV‐neoplasia (CV‐N) areas [4]. Here, we report four patients in whom CV‐SCAN was used to delineate lesion extent before ESD and discuss its potential role as an adjunctive technique for endoscopic management of UCAN.

2. Case Report

Four patients with ulcerative colitis in clinical remission underwent surveillance colonoscopy using CV‐SCAN at National Defense Medical College Hospital. The clinical characteristics of the patients are summarized in Table S1. Three patients had pancolitis, and one had left‐sided colitis. The median disease duration was 14 years (range, 10–26 years).

In all four patients, conventional white‐light imaging failed to clearly delineate the lesions. Two lesions appeared as subtle reddish mucosal abnormalities, whereas the remaining two could not be identified by conventional white‐light imaging alone. In these two cases, including Case 2 (Figure S1) and Case 3 (Figure 2), the lesions became apparent only after application of CV‐SCAN. In contrast, CV‐SCAN identified all lesions as CV‐N areas with distinct borders (Figures 1, 2, 3, 4 and Figure S1). Targeted biopsies obtained from the CV‐N‐positive areas confirmed UCAN in all cases.

FIGURE 2.

Endoscopic findings of ulcerative colitis–associated neoplasia (UCAN) in Case 3 at the initial endoscopic submucosal dissection (ESD). (A) White‐light imaging shows scarred mucosa, and UCAN is difficult to identify under conventional observation. (B) Crystal violet staining for colitis‐associated neoplasia (CV‐SCAN) at the same location demonstrates an extensive crystal violet‐neoplasia (CV‐N)–positive area. (C) Resected specimen showing the corresponding lesion extent. The tumor area is delineated by a yellow line. The horizontal margin is focally positive.

FIGURE 1.

Endoscopic and histopathological findings of ulcerative colitis–associated neoplasia in Case 1. (A) White‐light imaging shows a subtle flat reddish lesion with indistinct borders. The tumor is indicated by a white arrow. (B) Crystal violet staining for colitis‐associated neoplasia (CV‐SCAN) demonstrates a clearly demarcated crystal violet‐neoplasia (CV‐N) area. The CV‐N–positive region is outlined by white arrows. (C) Resected specimen showing the corresponding lesion extent. The tumor area is delineated by a yellow line.

FIGURE 3.

Endoscopic findings of ulcerative colitis–associated neoplasia (UCAN) in Case 3 during the second endoscopic submucosal dissection (ESD) for a positive horizontal margin after the initial resection. (A) White‐light imaging does not clearly demonstrate UCAN at the resection site. (B) Crystal violet staining for colitis‐associated neoplasia (CV‐SCAN) reveals a clearly demarcated intensely stained area corresponding to UCAN. (C) Resected specimen showing the corresponding lesion extent. The tumor area is delineated by a yellow line.

FIGURE 4.

Endoscopic findings of ulcerative colitis–associated neoplasia (UCAN) in Case 4. (A) White‐light imaging shows an indistinct UCAN. The area indicated by the white arrow corresponds to the site that was pathologically diagnosed as pT1a. (B) CV‐SCAN demonstrates an intensely stained area corresponding to the pT1a lesion. (C) Resected specimen showing the corresponding lesion extent. Areas of low‐grade dysplasia are delineated by a yellow line, and the pT1a component is delineated by a blue line. On the left side of the specimen, epithelial components are partially detached or degenerated, making histopathological evaluation difficult in that region. (D) CV‐SCAN image showing a broad intensely stained area adjacent to the site corresponding to the pT1a lesion. This finding suggested that the CV‐N‐positive area extended beyond the depressed component recognized by white‐light imaging.

Following histological confirmation, repeat colonoscopy with CV‐SCAN was performed to determine the lateral extent of each lesion before ESD. Additional biopsies obtained from the surrounding mucosa outside the CV‐N‐positive areas were negative for neoplasia. Based on these findings, ESD was performed in all four patients.

Histopathological examination revealed low‐grade dysplasia in two patients, mixed low‐ and high‐grade dysplasia in one patient, and well‐differentiated tubular adenocarcinoma with an associated low‐grade dysplastic component in one patient. In the patient with adenocarcinoma, the final pathological diagnosis was well‐differentiated tubular adenocarcinoma (6 × 5 mm; tub1> por2, tub2) with 620 µm submucosal invasion (pT1a), INFb, Ly0 (D2‐40), V1 (EVG), pHM0, and pVM0. Immunohistochemical analyses demonstrated p53 overexpression and increased Ki‐67 expression extending toward the mucosal surface, supporting the diagnosis of UCAN.

En bloc resection was achieved in all patients. In three patients, the extent of the CV‐N‐positive area corresponded closely with the pathological extent of neoplasia, and complete resection was achieved with negative horizontal margins. One patient had an extensive lesion in which a reddish depressed area was recognized on conventional white‐light imaging within the CV‐N‐positive mucosa (Figure 4). Targeted biopsy from the depressed lesion demonstrated high‐grade dysplasia, and adenocarcinoma could not be excluded. Because of the extensive lesion and the suspicion of carcinoma, surgical resection was recommended. However, the patient declined surgery because of advanced age and requested endoscopic treatment. After thorough discussion and informed consent, ESD was performed. One patient had a positive horizontal margin after the initial ESD. During repeat evaluation, residual neoplasia remained difficult to recognize under white‐light imaging but was visualized as a residual CV‐N‐positive area by CV‐SCAN (Figure 3). Additional ESD was subsequently performed, and histopathological examination confirmed residual UCAN within the CV‐N‐positive area. No residual neoplasia was identified after the second ESD procedure.

In the patient with adenocarcinoma, additional colectomy was recommended after multidisciplinary discussion because the final pathological examination revealed pT1a adenocarcinoma with submucosal invasion and venous invasion. However, the patient declined surgical treatment because of advanced age and opted for careful surveillance.

No procedure‐related adverse events occurred. All four patients have been followed for approximately 2 years after ESD, and no local recurrence has been observed. In the patient with adenocarcinoma, surveillance colonoscopy and computed tomography were performed at 6 months and 1 year after ESD. No local recurrence or metastatic disease has been detected during the follow‐up period.

3. Discussion

CV‐SCAN may facilitate delineation of UCAN before ESD, particularly in lesions with indistinct borders. In all four patients, lesion margins were difficult to define using conventional white‐light imaging, whereas CV‐SCAN identified CV‐N areas that corresponded closely with the pathological extent of neoplasia in most cases. However, this four‐case report does not support the use of CV‐SCAN as a standalone modality. Rather, CV‐SCAN should be considered an adjunctive technique for selected lesions in which delineation remains difficult despite conventional endoscopic assessment.

Although image‐enhanced endoscopy, magnifying endoscopy, and chromoendoscopy have substantially improved the detection and characterization of UCAN, accurate delineation of lateral tumor extent remains challenging because chronic inflammation, fibrosis, and mucosal distortion frequently obscure lesion borders. Recently, Kawabata et al. demonstrated that combining lesion‐border findings with pit‐pattern analysis improves discrimination between invasive and noninvasive UCAN [5]. Their study focused on predicting invasion depth, whereas the present report addresses the delineation of lateral tumor extent before ESD. Importantly, CV‐SCAN was used only to determine lesion boundaries and was not intended to assess invasion depth. In the patient with adenocarcinoma, treatment decisions were based on conventional endoscopic findings, targeted biopsy demonstrating high‐grade dysplasia with suspected carcinoma, and the patient's preference after declining surgery.

Histopathological examination confirmed characteristic features of UCAN, including bottom‐up morphogenesis, p53 overexpression, expansion of Ki‐67‐positive cells toward the mucosal surface, and reduced SATB2 expression, all of which have been reported in UCAN [2, 6, 7, 8]. Furthermore, the distribution of CV‐N‐positive areas closely corresponded with these pathological findings. Although the macroscopic appearance of resected specimens did not always clearly distinguish neoplastic from non‐neoplastic mucosa, our previous study demonstrated that CV‐SCAN‐positive mucosa shows characteristic molecular alterations associated with UC‐associated tumorigenesis [4]. Therefore, CV‐SCAN may reflect biological alterations of the superficial epithelium rather than gross morphological differences alone, explaining why lesion boundaries can be visualized despite subtle pathological surface findings.

The staining pattern observed in Case 2 differed from the other three cases, with the lesion appearing relatively less intensely stained than the surrounding inflamed mucosa. We speculate that this reflects differences in epithelial architecture, CV retention, and chronic inflammatory changes. Nevertheless, the lesion became recognizable only after CV‐SCAN during repeat examination performed for pre‐ESD negative biopsies. Because this lesion presented as clustered reddish elevations, it might also have been detected using conventional chromoendoscopy, image‐enhanced endoscopy, or magnifying endoscopy.

Case 3 illustrates both the potential utility and technical limitations of CV‐SCAN. Residual neoplasia after the initial ESD remained difficult to identify using white‐light imaging but was visualized as a residual CV‐N‐positive area. Histopathology confirmed residual UCAN within the stained area. To improve reproducibility, careful circumferential observation after CV application, repeated dye application around lesion borders, and patient repositioning to achieve uniform staining appear important, particularly in larger lesions.

CV has raised concerns regarding potential genotoxicity, and routine clinical use has declined with the widespread adoption of image‐enhanced endoscopy. Therefore, CV‐SCAN should not replace established modalities but may serve as an adjunctive technique for selected difficult lesions after standard evaluation. This study is limited by its small sample size, single‐center design, and lack of direct comparison with other endoscopic imaging techniques. Larger prospective studies are needed to clarify the incremental value of CV‐SCAN in contemporary UC surveillance.

Author Contributions

Akira Tomioka: conceptualization, investigation, endoscopic assessment, data curation, interpretation of findings, and writing – original draft. Nanoka Chiya: investigation, data curation, endoscopic assessment, and writing – review and editing. Ryota Hokari: supervision, interpretation of clinical findings, and writing – review and editing. Shomei Ryozawa: supervision and critical revision of the manuscript for important intellectual content. All authors reviewed and approved the final manuscript.

Funding

The authors have nothing to report.

Ethics Statement

This study was conducted in accordance with the Declaration of Helsinki and was approved by the Institutional Review Board of National Defense Medical College Hospital (approval numbers 4957 and 5015).

Consent

Written informed consent for publication of clinical details and images was obtained from all patients.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

TABLE S1: Clinical characteristics of patients with ulcerative colitis‐associated neoplasia (UCAN) by crystal violet staining for colitis‐associated neoplasia (CV‐SCAN).

All patients had no prior history of UCAN and were under surveillance with colonoscopy. Disease extent was defined according to the distribution of colonic involvement at the time of diagnosis.

FIGURE S1: Endoscopic findings of ulcerative colitis–associated neoplasia (UCAN) in Case 2. (A) Crystal violet staining for colitis‐associated neoplasia (CV‐SCAN) reveals a cluster of small elevated lesions showing reduced crystal violet staining, whereas the surrounding mucosa demonstrates slightly stronger staining. No corresponding white‐light image is available because the lesion was not identifiable by conventional white‐light imaging before CV‐SCAN. (B) Narrow‐band imaging (NBI) demonstrates the clustered small elevated lesions appearing brownish. (C) Resected specimen showing the corresponding lesion extent. The tumor area is delineated by a yellow line.

FIGURE S2: Histopathological findings of ulcerative colitis–associated neoplasia in Case 1. (A) Hematoxylin–eosin staining reveals low to high‐grade dysplasia. The boundary between neoplastic and non‐neoplastic epithelium is indicated by a black arrow, with the neoplastic area located on the left side. (B) Ki‐67 immunostaining shows prominent proliferative activity, particularly at the crypt base. The boundary between neoplastic and non‐neoplastic epithelium is indicated by a black arrow, with the neoplastic area located on the left side. (C) p53 immunostaining demonstrates overexpression. The boundary between neoplastic and non‐neoplastic epithelium is indicated by a black arrow, with the neoplastic area located on the left side. (D) SATB2 immunostaining shows reduced expression in the neoplastic epithelium. The boundary between neoplastic and non‐neoplastic epithelium is indicated by a black arrow, with the neoplastic area located on the left side.

FIGURE S3: Histopathological findings of ulcerative colitis–associated neoplasia in Case 2. (A) Hematoxylin–eosin staining reveals low‐grade dysplasia. The boundary between neoplastic and non‐neoplastic epithelium is indicated by a black arrow, with the neoplastic area located on the right side. (B) Ki‐67 immunostaining shows prominent proliferative activity extending to the surface epithelium. The boundary between neoplastic and non‐neoplastic epithelium is indicated by a black arrow, with the neoplastic area located on the right side. (C) p53 immunostaining demonstrates overexpression. The boundary between neoplastic and non‐neoplastic epithelium is indicated by a black arrow, with the neoplastic area located on the right side. (D) SATB2 immunostaining shows reduced expression in the neoplastic epithelium. The boundary between neoplastic and non‐neoplastic epithelium is indicated by a black arrow, with the neoplastic area located on the right side.

FIGURE S4: Histopathological findings of ulcerative colitis–associated neoplasia (UCAN) in Case 3 at the initial ESD. (A) Hematoxylin–eosin staining reveals low‐grade dysplasia. The boundary between neoplastic and non‐neoplastic epithelium is indicated by a black arrow, with the neoplastic area located on the right side. (B) Ki‐67 immunostaining shows prominent proliferative activity extending to the surface epithelium. The boundary between neoplastic and non‐neoplastic epithelium is indicated by a black arrow, with the neoplastic area located on the right side. (C) p53 immunostaining demonstrates overexpression. The boundary between neoplastic and non‐neoplastic epithelium is indicated by a black arrow, with the neoplastic area located on the right side. (D) SATB2 immunostaining shows reduced expression in the neoplastic epithelium. The boundary between neoplastic and non‐neoplastic epithelium is indicated by a black arrow, with the neoplastic area located on the right side.

FIGURE S5: Histopathological findings of ulcerative colitis–associated neoplasia (UCAN) in Case 3 during the second ESD for a positive horizontal margin after the initial resection. (A) Hematoxylin–eosin staining reveals low‐grade dysplasia. The boundary between neoplastic and non‐neoplastic epithelium is indicated by a black arrow, with the neoplastic area located on the right side. (B) Ki‐67 immunostaining shows prominent proliferative activity extending to the surface epithelium. The boundary between neoplastic and non‐neoplastic epithelium is indicated by a black arrow, with the neoplastic area located on the right side. (C) p53 immunostaining demonstrates overexpression. The boundary between neoplastic and non‐neoplastic epithelium is indicated by a black arrow, with the neoplastic area located on the right side. (D) SATB2 immunostaining shows reduced expression in the neoplastic epithelium. The boundary between neoplastic and non‐neoplastic epithelium is indicated by a black arrow, with the neoplastic area located on the right side.

FIGURE S6: Histopathological findings of ulcerative colitis–associated neoplasia (UCAN) in Case 4. (A) Hematoxylin–eosin staining reveals invasive carcinoma corresponding to pT1a (620 µm). (B) Ki‐67 immunostaining shows prominent proliferative activity extending to the surface epithelium. (C) p53 immunostaining demonstrates overexpression. (D) SATB2 immunostaining shows reduced expression in the neoplastic epithelium.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.