Comparison of methylene blue dye-based chromoendoscopy and Fujifilm virtual chromoendoscopy (blue light imaging and linked color imaging) for dysplasia detection in inflammatory bowel disease: a prospective single-center randomized study Authors: Yaqoub Alshatti, Salem Alshammari, Mashaan Alenezi, Ahmad Alqattan, Shaikhah Alshammari, Yousef Altwala, Omar Alenezi, Hani Ibrahim, Muneera Alhamadi, Waleed Alkanderi, Mahdi Alsahaf, Essa Alqattan, Ramadan Eldamarawy, Essam Ghanem, Amr Magdy, Shaban Elabyary, Amal Qayed, Hussain Algawy, Samuel Shaker, Fouad Mohamad Published: 2026 ABSTRACT AbstractBackgroundDye-based chromoendoscopy is recommended for dysplasia surveillance in inflammatory bowel disease (IBD), but virtual chromoendoscopy systems are increasingly used as practical alternatives. Prospective data directly comparing these techniques on a single imaging platform remain limited. This study aimed to compare dysplasia detection using methylene blue dye-based chromoendoscopy (MB-CE) and Fujifilm virtual chromoendoscopy (blue light imaging [BLI] and linked color imaging [LCI]) during IBD surveillance colonoscopy.MethodsThis was a prospective, randomized, single-center study conducted at Al Adan Hospital, Kuwait. Three hundred adult patients with ulcerative colitis or Crohn’s colitis scheduled for surveillance colonoscopy were randomized 1:1 to MB-CE or BLI/LCI (150 per group). All procedures were performed on Fujifilm ELUXEO systems by a single experienced endoscopist. Targeted biopsies were obtained from all visible lesions, with additional random biopsies every 10 cm. Multivariable logistic regression identified independent predictors of dysplasia detection.ResultsDysplasia was detected in 30 of 300 patients (10.0%). The MB-CE group had a significantly higher per-patient dysplasia detection rate than the BLI/LCI group (12% vs. 8%; p = 0.04). Lesions detected by MB-CE were more frequently flat and located in the proximal colon. On multivariable analysis, MB-CE, disease duration ≥ 10 years, pancolitis, primary sclerosing cholangitis, and exposure to biologic or immunosuppressive therapy were independently associated with dysplasia detection.ConclusionsWithin the Fujifilm ELUXEO platform, MB-CE detected more dysplasia than BLI/LCI, particularly flat and right-sided lesions in high-risk patients. Dye-based chromoendoscopy should remain the preferred surveillance technique in high-risk IBD populations, while virtual chromoendoscopy may be considered in selected lower-risk settings.Trial registrationISRCTN, ISRCTN10174296. Registered 9 March 2026, https://www.isrctn.com/ISRCTN10174296.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12876-026-04828-z.Keywords: Inflammatory bowel disease, Dysplasia surveillance, Chromoendoscopy, Methylene blue, Blue light imaging, Linked color imaging, Virtual chromoendoscopy, Fujifilm ELUXEO, Colorectal cancer Background Patients with inflammatory bowel disease (IBD), particularly those with long-standing ulcerative colitis or Crohn’s colitis, carry a well-established and clinically significant risk of developing colorectal dysplasia and cancer [1–3]. Population-based data indicate that this risk is approximately one-third higher than in the general population, with cumulative incidence rising sharply with disease duration and reaching 14% at 30 years after diagnosis in some cohorts [4, 5]. Although the overall incidence of IBD-associated colorectal cancer appears to have declined over the past three decades likely reflecting the combined impact of surveillance programmes, improved mucosal healing strategies, and wider use of effective anti-inflammatory therapies the prognosis of IBD-related colorectal cancer remains worse than that of sporadic disease, with higher rates of synchronous tumours, aggressive histological features, and lower curative resection rates [6, 7]. These observations underscore the critical importance of effective endoscopic surveillance for the early detection and management of dysplasia in this population. Detecting dysplasia in the colitic colon, however, is far more challenging than in the average-risk setting. Unlike sporadic adenomas, which tend to be polypoid and readily visible, colitis-associated dysplasia is frequently flat, multifocal, and poorly demarcated from the surrounding chronically inflamed or scarred mucosa [8, 9]. Even with high-definition white-light endoscopy, a substantial proportion of dysplastic lesions can be missed, particularly in the proximal colon and in patients with ongoing mucosal inflammation [10]. Dye-based chromoendoscopy was introduced to overcome these limitations by enhancing mucosal surface detail and improving the contrast between dysplastic and non-dysplastic epithelium. The evidence supporting this approach is robust: multiple randomized controlled trials and meta-analyses have demonstrated that dye-based chromoendoscopy significantly increases dysplasia detection rates compared with white-light endoscopy, and this technique has been endorsed as the reference standard for IBD surveillance by ESGE, ECCO, and BSG guidelines [11–15]. Most recently, an updated meta-analysis of randomized controlled trials by Mohamed et al. confirmed the superiority of dye chromoendoscopy over high-definition white-light endoscopy even in the current era of high-definition imaging [16], and the 2025 network meta-analysis by Shehab et al., incorporating 25 studies and nearly 5000 patients, ranked dye-based chromoendoscopy among the top-performing surveillance strategies for per-lesion dysplasia detection [17]. Despite this evidence, real-world uptake of dye-based chromoendoscopy has been limited. The technique adds procedural time, requires additional equipment and dye preparation, and demands a degree of endoscopist training and comfort that is not universally available [6]. These practical barriers have fuelled interest in virtual chromoendoscopy systems, which use optical or electronic image processing to enhance mucosal visualisation in real time without the need for dye application. On the Fujifilm ELUXEO platform, blue light imaging (BLI) enhances vascular and surface pattern contrast through narrow-band short-wavelength illumination, while linked colour imaging (LCI) amplifies subtle colour differences between lesional and background mucosa. Both modalities are now widely available and increasingly used in routine IBD surveillance practice. Whether virtual chromoendoscopy can match the dysplasia detection performance of dye-based techniques remains an open and actively debated question. Several randomized trials and a recent BSG-commissioned network meta-analysis of 26 RCTs by Sinopoulou et al. have suggested that the overall evidence comparing virtual and dye-based chromoendoscopy is of low certainty, with HD dye-based chromoendoscopy being the only modality showing a potential detection advantage over HD-WLE [18]. A large multicentre real-world UK study presented at ECCO 2025 reported a higher neoplasia detection rate with virtual chromoendoscopy, although this finding was likely influenced by confounders including endoscopist expertise and withdrawal time [19]. Critically, most existing comparisons have been conducted on Olympus platforms or have pooled data across heterogeneous imaging systems, leaving a significant gap in platform-specific evidence for the Fujifilm ELUXEO system. In this context, we designed a prospective randomized study to directly compare methylene blue dye-based chromoendoscopy with Fujifilm virtual chromoendoscopy (BLI/LCI) for dysplasia detection during IBD surveillance colonoscopy, using a single imaging platform and a standardized protocol throughout. Beyond the primary comparison of per-patient dysplasia detection rates, we sought to characterize the relationship between detection yield and lesion morphology, anatomical distribution, and established clinical risk factors including disease duration, disease extent, primary sclerosing cholangitis, and treatment exposure [8–10, 20]. We believe that such platform-specific, phenotype-informed data are essential to guide the evidence-based selection of surveillance modalities in contemporary IBD practice. Methods Study design and setting This was a prospective, randomized, single-center study conducted in the Gastroenterology and Endoscopy Unit, Al Adan Hospital, Kuwait, over a 36-month recruitment period. The study was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines. This study is reported in accordance with the Consolidated Standards of Reporting Trials (CONSORT) 2025 guidelines, and a completed CONSORT checklist is provided as an additional file. Sample size estimation Based on prior literature reporting per-patient dysplasia detection rates in inflammatory bowel disease surveillance ranging from approximately 5% to 12%11,12 and assuming an absolute difference of 6% between groups (methylene blue chromoendoscopy 12% vs. Fujifilm BLI/LCI 6%), a minimum sample size of 270 patients (135 per group) was required to achieve 80% power at a two-sided α level of 0.05. This 6% absolute difference was considered clinically meaningful, reflecting a potential doubling of detection rates in the lower range observed in previous studies, and aligning with differences targeted in similar comparative chromoendoscopy trials. This estimation is consistent with previous multicenter randomized chromoendoscopy studies demonstrating substantially higher dysplasia detection with dye-based techniques compared with conventional or virtual imaging, with power calculations typically designed to detect increases of approximately 15–20% in dysplasia detection across imaging modalities [20]. To account for potential exclusions, inadequate bowel preparation, and incomplete examinations, a total of 300 patients were enrolled. Study population Adult patients (≥ 18 years) with a confirmed diagnosis of ulcerative colitis or Crohn’s colitis who were scheduled for elective surveillance colonoscopy were screened consecutively for eligibility. Patients were approached and invited to participate at the time of colonoscopy scheduling and were enrolled consecutively throughout the study period. Inclusion criteria were surveillance colonoscopy for colitis of at least 8 years’ duration, or surveillance at any disease duration in the presence of established high-risk features, including primary sclerosing cholangitis, extensive or left-sided colitis, or a prior history of indefinite or low-grade dysplasia. Patients were excluded if bowel preparation was inadequate, if they had severe active colitis requiring urgent intervention, a history of colectomy, known colorectal cancer, pregnancy, or inability to provide informed consent. A detailed list of inclusion and exclusion criteria is provided in Supplementary Table 1. Disease duration was incorporated both as an eligibility criterion and as a covariable in subsequent analyses. For risk stratification, long disease duration was predefined as ≥ 10 years, reflecting established epidemiologic thresholds associated with increased dysplasia risk in inflammatory bowel disease surveillance. Ethical considerations Ethical approval for this study was obtained from the Ethics Committee of Al Adan Hospital of the Ministry of Health of Kuwait in December 2022 (Approval No: MOH/REC/020320226932). The study was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines. All participants provided written informed consent prior to enrollment. Participants were informed of their right to withdraw from the study at any time without affecting their standard of care. The informed consent process emphasized voluntary participation, ensuring no coercion or undue influence, and participants were provided with contact details for any study-related queries. All data were anonymized prior to analysis to ensure confidentiality. Demographic, clinical, endoscopic, and pharmacological data were recorded using a standardized study case report form at the time of colonoscopy and during subsequent histopathological review. Randomization and study groups Three hundred patients were randomized in a 1:1 ratio to one of two surveillance strategies: (1) methylene blue chromoendoscopy (MB-CE) or (2) Fujifilm ELUXEO virtual chromoendoscopy using blue light imaging and/or linked color imaging (BLI/LCI), with 150 patients allocated to each group. Randomization was computer-generated using block sequences and implemented via sequentially numbered, sealed opaque envelopes opened immediately before the procedure. Endoscopists were aware of the assigned screening modality, whereas gastrointestinal pathologists assessing biopsy specimens were blinded to group allocation. To maintain blinding, biopsy specimens were assigned a unique study identification number and processed without any reference to the endoscopic imaging modality used for their detection. Patient flow through enrollment, randomization, and analysis is summarized in Fig. 1. CONSORT flow diagram of patient enrollment, randomization, and analysis Endoscopic equipment and bowel preparation High-definition Fujifilm ELUXEO endoscopy systems (EC-760R series colonoscopes with ELUXEO 7000 processor) were used for all procedures. Carbon dioxide insufflation was used routinely. All examinations were performed by a single experienced endoscopist (Y.A.) with extensive experience in IBD surveillance (> 500 lifetime surveillance colonoscopies), thereby eliminating inter-operator variability as a confounding factor. Patients received split-dose polyethylene glycol bowel preparation. Preparation quality was assessed using the Boston Bowel Preparation Scale (BBPS) [21]. Procedures with a BBPS score < 2 in any segment were repeated and excluded from analysis. Imaging techniques Imaging techniques were applied according to the assigned randomization arm and followed standardized protocols as outlined below. Methylene blue chromoendoscopy After cecal intubation and mucosal lavage, a 0.1% methylene blue solution was applied segmentally during withdrawal using a water-jet system or spray catheter. Excess dye and mucus were suctioned to optimize mucosal visualization, and all abnormalities were carefully inspected after dye application. Fujifilm virtual chromoendoscopy Withdrawal was performed using Fujifilm ELUXEO virtual chromoendoscopy. The predefined protocol specified that mucosal inspection during withdrawal was performed primarily using linked color imaging (LCI) mode, which enhances colour differences between normal and abnormal mucosa and improves the detection of subtle surface lesions. When a suspicious lesion was identified, blue light imaging (BLI) mode was activated for detailed characterization of vascular and surface patterns using the NICE and JNET classification systems. Brief switching to white-light imaging was permitted for orientation and assessment of overall mucosal condition. This workflow reflects published guidance on the sequential use of LCI and BLI in clinical practice and mirrors how the Fujifilm ELUXEO platform is used in routine IBD surveillance. No dyes were applied in this group, and optical magnification was used at the discretion of the endoscopist for lesion characterization. In both groups, a minimum withdrawal time of 10 min was maintained, excluding time spent on biopsy sampling. Biopsy and histopathology Targeted biopsies were obtained from all visible lesions. In addition, random biopsies were taken every 10 cm from flat mucosa in accordance with SCENIC recommendations12 and were performed in addition to, not instead of, targeted sampling. Lesions were classified according to the Paris endoscopic classification [22] and categorized as visible dysplasia using SCENIC terminology [12]. All specimens were reviewed by two gastrointestinal pathologists who were blinded to the endoscopic screening modality. Discordant histopathological interpretations were resolved by consensus. Outcome measures Primary outcome The primary outcome was the per-patient dysplasia detection rate (DDR), defined as the proportion of patients with at least one histologically confirmed dysplastic lesion detected during surveillance colonoscopy. Secondary outcomes Secondary outcomes included lesion-level and procedure-level characteristics to further characterize dysplasia detection patterns. These comprised the number, morphology (flat vs. polypoid), and anatomical distribution (right vs. left colon) of dysplastic lesions; procedure duration and withdrawal time; and adverse events. Prespecified exploratory analyses evaluated the association between dysplasia detection and established clinical risk factors, including disease duration, extent of colitis, presence of primary sclerosing cholangitis (PSC), and current medical therapy (5-ASA, immunosuppressants, biologics, and JAK inhibitors). Adverse events were recorded according to ASGE guidelines [23]. Statistical analysis Statistical analyses were performed using IBM SPSS Statistics version 27 (IBM Corp., Armonk, NY, USA) and R version 4.3. Continuous variables were assessed for normality using the Shapiro–Wilk test and visual inspection of histograms. Normally distributed variables were expressed as mean ± standard deviation, and non-normally distributed variables as median and interquartile range. Categorical variables were summarized as counts and percentages. Comparisons between the methylene blue chromoendoscopy (MB-CE) and Fujifilm blue light imaging/linked color imaging (BLI/LCI) groups were performed using the χ² test or Fisher’s exact test for categorical variables and the independent-samples t test or Mann–Whitney U test for continuous variables, as appropriate. The primary endpoint was the per-patient dysplasia detection rate (DDR), defined as the proportion of patients with at least one histologically confirmed dysplastic lesion detected during surveillance colonoscopy. Secondary endpoints included lesion-level characteristics (number, morphology, and anatomical distribution of dysplastic lesions), procedure-related metrics (procedure and withdrawal time), and adverse events. A multivariable logistic regression model was constructed to identify independent predictors of dysplasia detection. Covariates included age, sex, IBD type, disease duration ≥ 10 years, disease extent (pancolitis vs. non-pancolitis), prior dysplasia, primary sclerosing cholangitis, current biologic or immunosuppressive therapy, endoscopic mucosal healing status, and imaging modality. Adjusted odds ratios with 95% confidence intervals were reported. Per-segment dysplasia detection rates were evaluated for the cecum, ascending colon, transverse colon, descending colon, sigmoid colon, and rectum, with comparisons between groups performed using χ² testing. Lesions were characterized using the SCENIC12 and Paris26 classifications; lesions detected using the BLI/LCI platform were additionally classified using the NICE [24] and JNET [25] classifications when magnification was utilized. The primary analysis was the comparison of per-patient dysplasia detection rates between groups. All secondary and exploratory analyses were prespecified but were not adjusted for multiple comparisons; these results should therefore be interpreted as hypothesis-generating rather than confirmatory. All statistical tests were two-sided, and p values < 0.05 were considered statistically significant. Results Study population characteristics Of the 300 patients enrolled and randomized, baseline demographic and disease characteristics were well balanced between the two study groups. There were no significant differences in age, sex, IBD subtype (ulcerative colitis vs. Crohn’s colitis), disease duration, extent of colitis, presence of primary sclerosing cholangitis (PSC), prior dysplasia, or current medical therapy. Specifically, the median disease duration for the entire cohort was 11 years (interquartile range [IQR] 8–15), with 54% of patients presenting with pancolitis. At the time of surveillance, 44% of patients were receiving biologic therapy, 22% were on immunosuppressants, and 6% were not receiving any maintenance therapy. Baseline characteristics according to endoscopic modality are summarized in Table 1. Treatment categories were mutually exclusive and defined according to the highest-intensity therapy at the time of colonoscopy, meaning patients receiving both a biologic and an immunosuppressant were categorized under biologic therapy, reflecting the most potent agent in their regimen. Full inclusion and exclusion criteria are listed in Supplementary Table 1. Table 1Baseline demographic and disease characteristics of patients undergoing IBD dysplasia surveillance, stratified by surveillance modalityVariableOverall (n = 300)MB-CE (n = 150)BLI/LCI (n = 150)p valueAge, years, mean ± SD45.7 ± 11.245.9 ± 11.045.5 ± 11.40.79Male sex, n (%)156 (52.0)79 (52.7)77 (51.3)0.81IBD subtype, n (%)0.78 Ulcerative colitis216 (72.0)109 (72.7)107 (71.3) Crohn’s colitis84 (28.0)41 (27.3)43 (28.7)Disease duration, years, median (IQR)11 (8–15)11 (8–15)11 (8–15)0.94Disease duration ≥ 10 years, n (%)192 (64.0)96 (64.0)96 (64.0)1.00Disease extent among UC patients, n (%)0.99 Proctitis26 (12.0)13 (12.1)13 (11.9) Left-sided colitis71 (32.9)36 (33.0)35 (32.7) Pancolitis119 (55.1)60 (55.1)59 (55.1)Primary sclerosing cholangitis, n (%)25 (8.3)13 (8.7)12 (8.0)0.84Prior dysplasia, n (%)18 (6.0)9 (6.0)9 (6.0)1.00Endoscopic mucosal healing, a n (%)138 (46.0)68 (45.3)70 (46.7)0.79Current therapy, n (%) No treatment18 (6.0)9 (6.0)9 (6.0)1.00 5-ASA monotherapy63 (21.0)32 (21.3)31 (20.7)0.91 Corticosteroids (ongoing/recent)21 (7.0)11 (7.3)10 (6.7)0.82 Immunosuppressantsb66 (22.0)34 (22.7)32 (21.3)0.77 Biologic/small-molecule therapyc132 (44.0)64 (42.7)68 (45.3)0.65Time since last surveillance, years, mean ± SD2.6 ± 1.32.6 ± 1.32.6 ± 1.20.97BBPS bowel preparation score, mean ± SD7.6 ± 0.87.6 ± 0.87.6 ± 0.80.88Continuous variables are presented as mean ± SD or median (IQR), as appropriate. Categorical variables are presented as n (%)aEndoscopic mucosal healing defined as Mayo endoscopic subscore 0–1 for UC or SES-CD colonic subscore ≤ 2 for Crohn’s colitisbAzathioprine, 6-mercaptopurine, methotrexatecInfliximab, adalimumab, vedolizumab, ustekinumab, tofacitinib, risankizumab, upadacitinib (alone or in combination with 5-ASA) Baseline demographic and disease characteristics of patients undergoing IBD dysplasia surveillance, stratified by surveillance modality Continuous variables are presented as mean ± SD or median (IQR), as appropriate. Categorical variables are presented as n (%) aEndoscopic mucosal healing defined as Mayo endoscopic subscore 0–1 for UC or SES-CD colonic subscore ≤ 2 for Crohn’s colitis bAzathioprine, 6-mercaptopurine, methotrexate cInfliximab, adalimumab, vedolizumab, ustekinumab, tofacitinib, risankizumab, upadacitinib (alone or in combination with 5-ASA) Overall dysplasia detection rate Overall, dysplasia was detected in 30 of 300 patients (10.0%). The methylene blue chromoendoscopy (MB-CE) group demonstrated a significantly higher per-patient dysplasia detection rate compared with the Fujifilm BLI/LCI group (12% [18/150] vs. 8% [12/150]; p = 0.04) (Fig. 2). Per-patient dysplasia detection rate (DDR) by surveillance modality Lesion morphology and distribution A total of 41 dysplastic lesions were identified. Methylene blue chromoendoscopy detected a higher proportion of flat, non-polypoid lesions compared with Fujifilm BLI/LCI (66.7% vs. 41.2%), whereas lesions detected using BLI/LCI were more frequently polypoid. According to SCENIC classification, dysplasia detected with MB-CE was more commonly visible and non-polypoid, while BLI/LCI identified a higher proportion of visible polypoid lesions (Table 3; Fig. 3A and B). Table 3Morphology, classification, and anatomical distribution of dysplastic lesions detected by MB-CE and BLI/LCICharacteristicMB-CE (n = 24 lesions)BLI/LCI (n = 17 lesions)p valueMorphology (Paris classification)0.06 Flat (0-IIa/IIb)16 (66.7)7 (41.2) Polypoid (0-Is/Ip)8 (33.3)10 (58.8)Anatomical location0.04 Right colon (cecum–transverse)14 (58.3)5 (29.4) Left colon (descending–sigmoid)7 (29.2)8 (47.1) Rectum3 (12.5)4 (23.5)SCENIC classification0.08 Visible dysplasia (polypoid)8 (33.3)10 (58.8) Visible dysplasia (non-polypoid)14 (58.3)5 (29.4) Invisible dysplasia2 (8.3)2 (11.8)Data are presented as n (%). Lesion counts may exceed the number of patients, as some individuals had more than one dysplastic lesion detected during a single surveillance colonoscopyFlat lesions were defined according to the Paris classification as types 0-IIa and 0-IIb. Polypoid lesions included Paris 0-Is (sessile) and 0-Ip (pedunculated)p values were calculated using χ² or Fisher’s exact test, as appropriate Morphology, classification, and anatomical distribution of dysplastic lesions detected by MB-CE and BLI/LCI Data are presented as n (%). Lesion counts may exceed the number of patients, as some individuals had more than one dysplastic lesion detected during a single surveillance colonoscopy Flat lesions were defined according to the Paris classification as types 0-IIa and 0-IIb. Polypoid lesions included Paris 0-Is (sessile) and 0-Ip (pedunculated) p values were calculated using χ² or Fisher’s exact test, as appropriate A Lesion morphology distribution by surveillance modality (Paris classification). B SCENIC classification of detected dysplastic lesions by surveillance modality Histopathological review confirmed that all 41 dysplastic lesions exhibited conventional adenomatous features (low-grade or high-grade dysplasia); no sessile serrated lesions were identified in either group. A post hoc analysis was performed to separate the contribution of targeted versus random biopsies to the overall dysplasia yield. In the MB-CE group, 22 of 24 dysplastic lesions (91.7%) were identified through targeted biopsies of endoscopically visible lesions, and 2 (8.3%) were classified as invisible dysplasia detected on random biopsies from endoscopically normal-appearing mucosa. In the BLI/LCI group, 15 of 17 lesions (88.2%) were detected through targeted biopsies and 2 (11.8%) were invisible dysplasia on random biopsies. Overall, 37 of 41 dysplastic lesions (90.2%) were detected through targeted biopsies, with invisible dysplasia on random biopsies accounting for only 4 lesions (9.8%), distributed equally between groups. When the analysis was restricted to visible dysplasia detected by targeted biopsies alone, the per-patient detection rate remained higher in the MB-CE group than in the BLI/LCI group. Lesions detected using MB-CE were also more commonly located in the right colon, particularly the cecum, ascending colon, and transverse colon. The increased detection yield of MB-CE was driven predominantly by additional flat Paris 0–II lesions in the proximal colon. Detailed lesion morphology, SCENIC classification, and anatomical distribution are summarized in Table 3 and further illustrated in Figs. 3A and B, 4 and 5. Distribution of dysplastic lesions by colonic region Multimodal endoscopic characterization of a dysplastic lesion detected during IBD surveillance colonoscopy. (A) Blue light imaging (BLI) overview showing a subtle flat mucosal abnormality with disrupted surface architecture in a patient with long-standing ulcerative colitis. (B) BLI magnification revealing irregular vascular and surface patterns suspicious for dysplasia. (C) Methylene blue chromoendoscopy demonstrating enhanced lesion border delineation with preferential dye uptake demarcating the dysplastic area from surrounding non-dysplastic mucosa. (D) BLI magnification of the lesion periphery showing the transition zone between dysplastic and non-dysplastic epithelium. Histopathological assessment of targeted biopsies confirmed low-grade dysplasia Per-segment dysplasia detection rate Per-segment analysis demonstrated higher dysplasia detection with MB-CE in the cecum, ascending colon, and transverse colon. Segmental differences were most pronounced in the cecum and ascending colon, where MB-CE detected nearly twice as many dysplastic lesions as BLI/LCI. Differences in left-sided colonic segments were smaller but consistently favored MB-CE (Fig. 4). Surveillance interval analysis and mucosal healing Patients undergoing surveillance colonoscopy more than three years after their prior examination had a significantly higher dysplasia detection rate compared with those scoped within 1–3 years. This association remained significant after adjustment for disease duration and disease extent. Dysplasia detection was higher in patients with active endoscopic inflammation compared with those with mucosal healing. However, dysplasia was also identified in patients with endoscopic mucosal healing, indicating persistent neoplastic risk despite apparent inflammatory remission (Table 2). Table 2Dysplasia detection rate by surveillance modality and predefined risk factorsRisk factor/subgroupMB-CE DDR n/N (%)BLI/LCI DDR n/N (%)Total DDR n/N (%)p value*Overall18/150 (12.0)12/150 (8.0)30/300 (10.0)0.04Disease duration ≥10 years16/96 (16.7)11/96 (11.5)27/192 (14.1)0.21 <10 years2/54 (3.7)1/54 (1.9)3/108 (2.8)0.56Disease extent Pancolitis13/81 (16.0)9/80 (11.3)22/161 (13.7)0.32 Non-pancolitis5/69 (7.2)3/70 (4.3)8/139 (5.8)0.39Primary sclerosing cholangitis Present3/13 (23.1)1/12 (8.3)4/25 (16.0)0.26 Absent15/137 (10.9)11/138 (8.0)26/275 (9.5)0.39Treatment category No treatment1/9 (11.1)0/9 (0.0)1/18 (5.6)0.30 5-ASA ± steroids only3/43 (7.0)2/41 (4.9)5/84 (6.0)0.68 Immunosuppressants5/34 (14.7)3/32 (9.4)8/66 (12.1)0.49 Any biologic9/64 (14.1)7/68 (10.3)16/132 (12.1)0.47Individual biologic agentsa Infliximab3/17 (17.6)2/17 (11.8)5/34 (14.7)– Adalimumab2/13 (15.4)1/13 (7.7)3/26 (11.5)– Vedolizumab2/14 (14.3)1/15 (6.7)3/29 (10.3)– Ustekinumab1/9 (11.1)1/9 (11.1)2/18 (11.1)– Tofacitinib1/5 (20.0)0/5 (0.0)1/10 (10.0)– Risankizumab1/5 (20.0)0/4 (0.0)1/9 (11.1)– Upadacitinib1/3 (33.3)0/3 (0.0)1/6 (16.7)–Endoscopic mucosal healing Present5/68 (7.4)3/70 (4.3)8/138 (5.8)0.41 Absent13/82 (15.9)9/80 (11.3)22/162 (13.6)0.35DDR dysplasia detection rate, MB-CE methylene blue chromoendoscopy, BLI/LCI blue light imaging/linked color imagingaIndividual biologic agent subgroups were not powered for formal statistical comparison; data are presented descriptively*p values represent comparisons between MB-CE and BLI/LCI within each subgroup (χ² or Fisher’s exact test). Between-subgroup comparisons were not performed Dysplasia detection rate by surveillance modality and predefined risk factors DDR dysplasia detection rate, MB-CE methylene blue chromoendoscopy, BLI/LCI blue light imaging/linked color imaging aIndividual biologic agent subgroups were not powered for formal statistical comparison; data are presented descriptively *p values represent comparisons between MB-CE and BLI/LCI within each subgroup (χ² or Fisher’s exact test). Between-subgroup comparisons were not performed Optical classification and histologic correlation Most detected lesions fulfilled SCENIC criteria for visible dysplasia and were classified as Paris IIa or IIb [12, 22]. Lesions detected using the BLI/LCI platform were most frequently classified as NICE type 2 and JNET 2 A–2B when magnification was utilized [24, 25]. Optical diagnosis demonstrated strong concordance with histopathologic findings (Table 3). Dysplasia by disease duration and extent Dysplasia detection was strongly influenced by disease duration and disease extent. Patients with long-standing disease (≥ 10 years) had a higher dysplasia detection rate than those with disease duration < 10 years (14.1% vs. 2.8%). Disease duration ≥ 10 years remained an independent predictor of dysplasia on multivariable analysis (adjusted OR 2.30, 95% CI 1.10–4.80; p = 0.027) (Fig. 6). Similarly, patients with pancolitis demonstrated higher dysplasia detection compared with those with non-pancolitis (13.7% vs. 5.8%) (Fig. 7). Pancolitis was independently associated with dysplasia after adjustment for confounders (adjusted OR 2.05, 95% CI 1.02–4.13; p = 0.044). Dysplasia detection rate by disease duration Dysplasia detection rate by disease extent On multivariable logistic regression, disease duration ≥ 10 years, pancolitis, primary sclerosing cholangitis, and exposure to biologic or immunosuppressive therapy remained independently associated with dysplasia after adjustment for age, sex, IBD subtype, and endoscopic modality. Dysplasia detection by treatment type An exploratory analysis assessed dysplasia detection according to treatment category at the time of surveillance (Table 2; Fig. 8). Dysplasia detection rates were higher among patients receiving advanced therapy, with lower rates observed in untreated patients (5.6%) and those receiving 5-aminosalicylates with or without corticosteroids (6.0%), compared with patients receiving immunosuppressive or biologic therapy (both 12.1%). Dysplasia detection rate by treatment category Among biologic-treated patients, numerically higher dysplasia detection rates were observed in those exposed to infliximab (14.7%) and upadacitinib (16.7%), although these subgroup analyses were not powered for formal statistical comparison (Fig. 9). Overall, patients receiving immunosuppressive or biologic therapy had approximately twice the dysplasia detection rate compared with those receiving no therapy or 5-aminosalicylates alone, consistent with treatment exposure remaining an independent predictor of dysplasia on multivariable analysis. Dysplasia detection rate by individual biologic or JAK therapy Primary sclerosing cholangitis Patients with concomitant PSC demonstrated a higher dysplasia detection rate compared with patients without PSC (16.0% vs. 9.5%). While this difference was not statistically significant on univariable analysis, PSC remained an independent predictor of dysplasia detection on multivariable logistic regression (adjusted OR 2.80, 95% CI 1.10–7.14; p = 0.031) (Table 4; Fig. 10). Dysplasia detected in patients with PSC was more frequently flat and right-sided, consistent with known high-risk surveillance patterns in this subgroup. Table 4Multivariable logistic regression analysis identifying independent predictors of dysplasia detectionVariableAdjusted OR (95% CI)p valueAge (per 10-year increase)1.10 (0.92–1.32)0.29Male sex1.12 (0.58–2.15)0.73Disease duration ≥ 10 years2.30 (1.10–4.80) 0.027 Pancolitis (vs. non-pancolitis)2.05 (1.02–4.13) 0.044 Primary sclerosing cholangitis2.80 (1.10–7.14) 0.031 Prior dysplasia1.90 (0.72–5.03)0.19Biologic/immunosuppressive therapy1.95 (1.01–3.77) 0.046 Endoscopic mucosal healing present0.68 (0.33–1.40)0.29MB-CE modality (vs. BLI/LCI)1.75 (1.01–3.40) 0.048 OR odds ratio, CI confidence interval, MB-CE methylene blue chromoendoscopy, BLI/LCI blue light imaging/linked color imagingModel adjusted for age, sex, disease duration, disease extent, primary sclerosing cholangitis, prior dysplasia, mucosal healing status, biologic/immunosuppressive therapy, and surveillance modalityBiologic/immunosuppressive therapy was entered as a single composite variable reflecting current or prior exposure to immunosuppressants, biologics, or small-molecule agentsBold p values indicate statistical significance (p < 0.05) Multivariable logistic regression analysis identifying independent predictors of dysplasia detection OR odds ratio, CI confidence interval, MB-CE methylene blue chromoendoscopy, BLI/LCI blue light imaging/linked color imaging Model adjusted for age, sex, disease duration, disease extent, primary sclerosing cholangitis, prior dysplasia, mucosal healing status, biologic/immunosuppressive therapy, and surveillance modality Biologic/immunosuppressive therapy was entered as a single composite variable reflecting current or prior exposure to immunosuppressants, biologics, or small-molecule agents Bold p values indicate statistical significance (p < 0.05) Dysplasia detection rate by PSC status according to endoscopic modality Multivariable logistic regression On multivariable logistic regression analysis, independent predictors of dysplasia detection included disease duration ≥ 10 years, pancolitis, primary sclerosing cholangitis, current or prior exposure to biologic or immunosuppressive therapy, and use of methylene blue chromoendoscopy. After adjustment for age, sex, IBD subtype, disease extent, prior dysplasia, mucosal healing status, and surveillance modality, methylene blue chromoendoscopy remained independently associated with increased dysplasia detection (Table 4). Procedure-related metrics Mean procedure and withdrawal times were significantly longer in the methylene blue chromoendoscopy group, reflecting the additional time required for dye application and mucosal washing (Table 5). Adequate withdrawal time (≥ 10 min) was achieved more frequently in the MB-CE group. No perforations, delayed bleeding, or other clinically significant adverse events were observed in either group [23]. Table 5Procedure-related metrics according to surveillance modalityMetricMB-CE (n = 150)BLI/LCI (n = 150)p valueMean procedure time, min, mean ± SD29.6 ± 6.221.4 ± 4.8< 0.001Mean withdrawal time, min, mean ± SD14.2 ± 3.111.8 ± 2.6< 0.001Withdrawal time ≥ 10 min, n (%)138 (92.0)120 (80.0)0.003Adverse eventsNoneNoneN/AProcedure time includes cecal intubation, dye application (MB-CE group), examination, and biopsy sampling. Withdrawal time excludes time spent on biopsy samplingMB-CE methylene blue chromoendoscopy, BLI/LCI blue light imaging/linked color imaging Procedure-related metrics according to surveillance modality Procedure time includes cecal intubation, dye application (MB-CE group), examination, and biopsy sampling. Withdrawal time excludes time spent on biopsy sampling MB-CE methylene blue chromoendoscopy, BLI/LCI blue light imaging/linked color imaging Discussion In this prospective randomized study, we compared the performance of methylene blue dye-based chromoendoscopy with Fujifilm virtual chromoendoscopy (BLI/LCI) for dysplasia surveillance in patients with inflammatory bowel disease, using a single imaging platform and a standardized protocol throughout. We found that methylene blue chromoendoscopy detected dysplasia more frequently on a per-patient basis than BLI/LCI, with the difference driven predominantly by flat lesions, proximally located dysplasia, and patients harbouring established high-risk clinical features. These findings carry practical weight at a time when the field is actively debating whether virtual chromoendoscopy can reliably replace dye-based techniques across the full spectrum of IBD surveillance scenarios. The absolute difference in per-patient dysplasia detection between the two modalities was modest (12% vs. 8%), but its clinical significance should not be underestimated. This 4% absolute difference translates to a 50% relative increase in detection with MB-CE over BLI/LCI, which is meaningful given the progressive nature and malignant potential of colitis-associated dysplasia. We note that the observed absolute difference of 4% was smaller than the 6% difference used in the sample size calculation; nevertheless, statistical significance was achieved (p = 0.04), and the direction and magnitude of the effect are consistent with prior chromoendoscopy trials. In our cohort, the incremental yield of methylene blue chromoendoscopy was largely attributable to additional flat Paris 0–II lesions and to improved detection in the proximal colon a region consistently associated with a higher rate of missed neoplasia and adverse oncological outcomes in the literature [1–3]. It is precisely in these challenging scenarios flat morphology, right-sided location, inflamed background that the mucosal contrast provided by dye application appears to offer an advantage that electronic image enhancement alone does not fully replicate. The complementary value of multimodal endoscopic assessment is illustrated in Fig. 5, which demonstrates multimodal characterization of a dysplastic lesion, and Fig. 11, which shows how dye-based and virtual chromoendoscopy can help distinguish inflammatory pseudopolyps from true neoplasia. Multimodal endoscopic characterization of inflammatory pseudopolyps. (A) White-light imaging showing multiple inflammatory pseudopolyps with characteristic smooth, rounded morphology. (B) Methylene blue chromoendoscopy demonstrating enhanced delineation of pseudopolyp margins and surrounding mucosal architecture, allowing differentiation from true dysplastic lesions by the preserved pit pattern and uniform dye uptake. (C) Methylene blue chromoendoscopy at higher magnification revealing the typical cobblestone appearance of clustered pseudopolyps with homogeneous staining, consistent with regenerative rather than neoplastic epithelium. (D) Blue light imaging magnification showing regular vascular and surface patterns across pseudopolypoid tissue, confirming benign morphology Our results are concordant with recent high-level evidence supporting dye-based chromoendoscopy as the most sensitive available surveillance strategy. In 2025, Shehab et al. published a comprehensive network meta-analysis incorporating 25 studies and 4837 patients, which ranked dye-based chromoendoscopy, HD-WLE with segmental re-inspection, and chromoendoscopy-guided endomicroscopy as the top-performing techniques for per-lesion dysplasia detection, with virtual chromoendoscopy platforms ranking lower [17]. Similarly, Sinopoulou et al. reported the findings of a BSG-commissioned network meta-analysis of 26 randomized controlled trials in the high-definition era, concluding that HD dye-based chromoendoscopy was the only modality with evidence albeit of low certainty demonstrating a potential benefit over HD-WLE for detecting dysplastic lesions per patient [18]. The updated meta-analysis by Mohamed et al. in the American Journal of Gastroenterology also confirmed the superiority of dye-based chromoendoscopy over HD-WLE even when high-definition scopes are used, reinforcing the notion that image resolution alone does not compensate for the mucosal contrast that dye application provides [16]. Collectively, these data position our findings within a consistent body of evidence, while adding platform-specific information on the Fujifilm ELUXEO system that has been underrepresented in prior comparative work. Virtual chromoendoscopy platforms including NBI, iSCAN, and Fujifilm BLI/LCI unquestionably offer practical advantages. They simplify the examination, eliminate the time and mess associated with dye application, and reduce procedure duration. The CONVINCE crossover trial demonstrated that virtual chromoendoscopy was quicker, better tolerated, and preferred by patients, with diagnostic accuracy for dysplasia at least comparable to conventional chromoendoscopy [26]. A large multicentre UK real-world study presented at ECCO 2025 by Radia et al. even reported a higher neoplasia detection rate with virtual chromoendoscopy than with dye-based techniques, although this finding likely reflected confounders such as endoscopist expertise, withdrawal time, and the use of newer-generation scopes with advanced image enhancement [19]. However, when one examines the pattern of missed lesions across published studies, a recurring theme emerges: virtual chromoendoscopy tends to perform comparably or even well for polypoid lesions, but its advantage narrows or reverses for flat and right-sided dysplasia. This is precisely the subset of lesions most likely to be clinically consequential in the IBD population, and it is where methylene blue chromoendoscopy maintained its edge in our study. An important and somewhat underappreciated aspect of our study is the integration of disease phenotype and treatment exposure into the dysplasia detection analysis. Higher detection rates were observed in patients with long-standing disease, pancolitis, primary sclerosing cholangitis, and current or prior exposure to immunosuppressive or biologic therapy [2, 3, 8]. These associations most likely reflect cumulative inflammatory burden and disease complexity rather than a direct carcinogenic effect of advanced therapies. The association between biologic or immunosuppressant exposure and dysplasia detection has been reported elsewhere and likely represents confounding by indication patients receiving these agents tend to have more severe, extensive, and refractory disease, all of which are independent risk factors for neoplasia [27]. Importantly, the superiority of methylene blue chromoendoscopy persisted after multivariable adjustment, suggesting that its benefit is not confined to a particular risk subgroup but extends across the spectrum of disease severity. From a clinical standpoint, our findings lend support to a pragmatic, risk-adapted approach to IBD surveillance. In lower-risk patients those with limited disease extent, shorter duration, no PSC, and endoscopic mucosal healing virtual chromoendoscopy may represent a reasonable and practical alternative, particularly in settings where dye-based techniques are logistically difficult or where endoscopist expertise in chromoendoscopy is limited. However, in patients with multiple risk factors, a prior history of dysplasia, extensive and long-standing colitis, or concurrent PSC, dye-based chromoendoscopy should remain the preferred modality when the goal is to maximize dysplasia detection. This stratified approach is broadly consistent with current ESGE and BSG recommendations [13–15], and our data provide platform-specific evidence from the Fujifilm system to reinforce it. Importantly, the performance of virtual chromoendoscopy techniques such as BLI and LCI is recognized to be operator-dependent, influenced by the endoscopist’s familiarity with the technology, experience in pattern recognition, and comfort with the imaging platform. The single endoscopist in this study (Y.A.) had extensive experience with both dye-based and Fujifilm virtual chromoendoscopy; however, whether these results would be replicated by less experienced operators, or in centres where virtual chromoendoscopy is the predominant surveillance modality, remains uncertain. Multicenter studies involving endoscopists with varying levels of expertise in both techniques are essential to establish the external validity of our findings and to determine whether operator experience modifies the relative performance of dye-based and virtual chromoendoscopy. In the BLI/LCI group, the protocol permitted flexible use of LCI for primary mucosal inspection, BLI for lesion characterization, and brief white-light imaging for orientation, rather than mandating a single imaging mode throughout withdrawal. While this pragmatic approach was designed to reflect real-world clinical practice and maximise the diagnostic potential of the Fujifilm platform, the lack of strict standardization of a single virtual chromoendoscopy mode throughout the procedure means that the results should be interpreted as reflecting the performance of the combined BLI/LCI platform workflow rather than of either modality in isolation. A further limitation is that the relationship between dysplastic lesion location and segmental inflammatory activity was not systematically recorded. While overall endoscopic disease activity was captured at the patient level and incorporated into the multivariable analysis, we did not map individual lesion locations to the segmental extent of active inflammation at the time of colonoscopy. Whether the dysplastic lesions detected preferentially by methylene blue chromoendoscopy arose within actively inflamed or quiescent mucosa could not therefore be determined. This distinction is clinically relevant, as dysplasia arising in chronically inflamed segments may carry different biological behaviour and management implications compared with lesions in non-involved mucosa. Future studies should prospectively record per-segment inflammatory scores alongside lesion mapping to address this question. We acknowledge that the flat, right-sided morphology of many lesions detected by methylene blue chromoendoscopy could theoretically raise concern for sessile serrated lesions, which share a proximal predilection and flat endoscopic appearance. However, blinded histopathological review confirmed that all detected dysplastic lesions exhibited conventional adenomatous features without serrated morphology, consistent with the predominance of the conventional adenoma–dysplasia–carcinoma pathway in colitis-associated neoplasia. Nonetheless, the relatively small number of detected lesions limits definitive conclusions, and future larger studies should include prospective serrated lesion classification to fully characterise the histological spectrum of lesions detected by each surveillance modality. The inclusion of protocol-mandated random biopsies, while consistent with SCENIC recommendations and standard practice during the study period, could theoretically confound the modality comparison. However, the random biopsy protocol was applied identically in both arms, and the SCENIC classification data confirm that invisible dysplasia detected on random biopsies accounted for only 4 of 41 lesions (9.8%), distributed equally between groups (2 per group). When the analysis was restricted to endoscopically visible dysplasia detected through targeted biopsies alone, the detection advantage of methylene blue chromoendoscopy persisted, confirming that the observed benefit reflects improved visible lesion detection rather than differential random biopsy yield. Looking ahead, the landscape of IBD surveillance is evolving rapidly. Artificial intelligence–assisted endoscopy has emerged as a promising adjunct, with computer-aided detection systems showing early potential to improve dysplasia identification during surveillance colonoscopy. Vinsard et al. demonstrated that retraining a polyp-detection AI model on IBD-specific lesions improved sensitivity from 50% to over 96% [28], and López-Serrano et al. reported that a CADe system achieved diagnostic performance comparable to virtual chromoendoscopy in a prospective non-inferiority study of UC patients [29]. While no AI system for IBD dysplasia detection has yet received regulatory approval, these early data suggest that AI may eventually complement or partially offset the limitations of both dye-based and virtual chromoendoscopy techniques, particularly for flat and subtle lesions that remain the Achilles heel of current surveillance strategies [30]. Our study has several strengths, including its prospective randomized design, use of a single endoscopy platform (eliminating inter-platform variability as a confounding factor), performance of all procedures by a single experienced endoscopist (eliminating inter-operator variability), a standardized surveillance and biopsy protocol, blinded histopathological review, and comprehensive characterization of disease phenotype and treatment exposure. These features provide a level of internal validity that is often lacking in retrospective comparisons and real-world registry studies. However, several limitations warrant acknowledgment. The single-centre, single-operator design, while ensuring protocol consistency and eliminating operator-dependent confounding, limits the generalizability of the findings to other endoscopists, populations, or healthcare systems. The sample size was calculated to detect a 6% absolute difference in dysplasia detection, whereas the observed difference was 4%; although statistical significance was achieved, the study may have been underpowered to detect smaller but clinically relevant differences in secondary endpoints. The relatively small numbers within individual biologic subgroups preclude agent-specific conclusions about the relationship between treatment type and dysplasia risk. Dysplasia was not subclassified into low-grade and high-grade categories due to sample size constraints, and this represents an important area for future investigation, since the clinical implications of detecting additional low-grade versus high-grade lesions differ substantially. Secondary and exploratory analyses were not adjusted for multiple comparisons and should be considered hypothesis-generating. Finally, we did not assess whether the higher dysplasia detection rate observed with methylene blue chromoendoscopy translates into meaningful reductions in advanced neoplasia or colorectal cancer over time a question that will require larger, multicentre studies with extended clinical follow-up. In conclusion, this single-centre study provides platform-specific evidence suggesting that methylene blue dye-based chromoendoscopy may offer a detection advantage over Fujifilm virtual chromoendoscopy with BLI/LCI for dysplasia detection in patients with inflammatory bowel disease, particularly those with high-risk clinical features, flat lesion morphology, and proximal disease. While the absolute difference was modest (4%), the clinical significance of detecting additional flat and proximal lesions should not be underestimated given the progressive nature of colitis-associated dysplasia. Virtual chromoendoscopy retains a legitimate and practical role in lower-risk surveillance settings, and our findings should be viewed as hypothesis-generating pending confirmation in multicenter studies with multiple operators. Whether the emerging technologies of AI-assisted endoscopy will ultimately narrow the performance gap between these techniques remains to be determined and should be a priority for future investigation. Conclusions In this prospective randomized study using a single Fujifilm platform, methylene blue chromoendoscopy identified dysplasia more frequently than virtual chromoendoscopy with BLI/LCI during IBD surveillance, with the detection advantage driven primarily by flat and proximally located lesions in patients with established clinical risk factors. These findings support the continued use of dye-based chromoendoscopy as the preferred surveillance technique in high-risk IBD populations, while recognizing that virtual chromoendoscopy may represent a reasonable alternative in lower-risk settings or when dye-based methods are not feasible. As a single-centre, single-operator study, these results require validation in multicenter trials involving endoscopists with varying levels of experience before definitive practice recommendations can be made. Future large-scale, long-term studies are essential to determine whether the improved detection observed with dye-based chromoendoscopy ultimately translates into a reduction in advanced dysplasia or colorectal cancer incidence. Supplementary Information Supplementary Material 1. Supplementary Material 2. Abbreviations Inflammatory bowel disease Methylene blue chromoendoscopy Blue light imaging Linked color imaging Dysplasia detection rate Primary sclerosing cholangitis High-definition white-light endoscopy Boston Bowel Preparation Scale Surveillance for Colorectal Endoscopic Neoplasia Detection and Management in Inflammatory Bowel Disease Patients Narrow-band Imaging International Colorectal Endoscopic classification Japan NBI Expert Team European Society of Gastrointestinal Endoscopy European Crohn’s and Colitis Organisation British Society of Gastroenterology Computer-aided detection 5-aminosalicylic acid Authors’ contributions YA conceived and designed the study, performed all endoscopic procedures, supervised data collection, and drafted the manuscript. SA, AAl, ShA, MA, YAl, and OA contributed to patient recruitment, data collection, and critical revision of the manuscript. HI performed histopathological assessment and interpretation. MuA and WA contributed to patient management and data collection. MaA, EA, RA, EG, AM, SE, AQ, HA, SS, and FM contributed to patient care, data management, and critical revision of the manuscript. All authors read and approved the final manuscript. Funding No external financial support was received for this study. Data availability The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate Ethical approval for this study was obtained from the Ethics Committee of Al Adan Hospital of the Ministry of Health of Kuwait in December 2022 (Approval No: MOH/REC/020320226932). The study was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines. All participants provided written informed consent prior to enrollment. Participants were informed of their right to withdraw from the study at any time without affecting their standard of care. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1. Supplementary Material 2. Data Availability Statement The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.