Linked color imaging (LCI), a dye-free form of virtual chromoendoscopy, detected precancerous and cancerous colorectal lesions about as often as traditional dye-chromoendoscopy (DCE) during surveillance colonoscopy in patients with long-standing inflammatory bowel disease (IBD) — and took less time to perform — according to a retrospective cohort study published in the Journal of Crohn's and Colitis.
Among 311 adults, DCE detected a mean of 0.116 neoplastic lesions per colonoscopy compared with 0.085 for LCI (P = .472), and the two techniques flagged at least one lesion in a similar share of procedures (8.03% vs. 7.53%; odds ratio [OR], 1.07; 95% CI, 0.45-2.54; P = .874). After propensity-score matching, imaging technique was not associated with the number of lesions detected (incidence rate ratio [IRR], 0.77; 95% CI, 0.34-1.78; P = .543).
Patients with long-standing colonic IBD carry an elevated risk of colorectal cancer, and guidelines recommend enhanced-imaging surveillance colonoscopy to catch dysplasia — abnormal tissue changes that can progress to cancer. DCE has been the preferred technique under the SCENIC consensus, but its uptake is low, the authors noted, largely because applying dye lengthens the procedure. LCI enhances mucosal and vascular contrast through image processing without dye, and the authors described their analysis as, in their words, the “first study to specifically evaluate LCI” for surveillance in colonic IBD.
The study was led by Roberto Gabbiadini and corresponding author Alessandro Armuzzi, MD, PhD, Full Professor of Gastroenterology at Humanitas University and head of the IBD Unit at IRCCS Humanitas Research Hospital in Rozzano, Milan, Italy.
The single-center team reviewed surveillance colonoscopies performed between June 2018 and December 2024 in adults with long-standing (eight or more years) ulcerative colitis or colonic Crohn's disease. Of 404 patients screened, 93 were excluded for inadequate bowel preparation or endoscopically active disease, leaving 311 for analysis — 112 examined with DCE using 0.04% methylene blue and 199 with LCI, all on high-definition colonoscopes. The primary outcome was the mean number of neoplastic lesions detected per colonoscopy; the secondary outcome was the proportion of colonoscopies detecting at least one lesion.
Overall, 30 neoplastic lesions were identified, all endoscopically visible, with no significant difference between techniques on either outcome, and the result held after adjustment for baseline differences between the groups.
Patient factors mattered more than the imaging technique, the authors reported. In multivariable analysis, a first-degree family history of colorectal cancer (IRR, 3.64; 95% CI, 1.34-9.93; P = .012) and longer disease duration (IRR, 1.05 per year; 95% CI, 1.01-1.10; P = .009) were independently associated with more lesions detected. Within high-definition systems and structured surveillance, the authors wrote, the choice of chromoendoscopy modality “appears less critical than established risk factors” such as disease duration and family history.
The techniques differed on time: colonoscope withdrawal averaged 14.2 minutes with LCI versus 17.5 minutes with DCE (P < .0001). Because LCI requires no dye, the authors suggested it may be easier to fold into routine practice while maintaining comparable diagnostic performance, and concluded that LCI may be adopted as an alternative to DCE for dysplasia surveillance in IBD.
The authors cautioned against overreading the equivalence. The study was retrospective, single-center and conducted by experienced endoscopists, limiting generalizability to lower-volume settings. The relatively small number of neoplastic lesions — reflecting a contemporary, lower-risk surveillance population — means the analysis “may be underpowered to detect small differences” between the techniques, they wrote, and they called for adequately powered prospective or randomized trials to confirm the findings.
The study reported no external funding. Several authors disclosed financial relationships with pharmaceutical and device makers; notably, co-authors Cesare Hassan and Alessandro Repici reported consulting relationships with Fujifilm, which manufactures the LCI system evaluated, and senior author Alessandro Armuzzi reported consulting, speaker and/or research ties to numerous drugmakers including AbbVie, Bristol-Myers Squibb, Eli Lilly, Janssen, MSD, Pfizer and Takeda. The remaining authors reported no competing interests.