Modular laser-based endoluminal ablation of the gastrointestinal tract: in vivo dose–effect evaluation and predictive numerical model

Background: Endoscopic submucosal dissection allows for “en bloc” removal of early gastrointestinal neoplasms. However, it is technically demanding and time-consuming. Alternatives could rely on energy-based techniques. We aimed to evaluate a predictive numerical model of thermal damage to preoperatively define optimal laser settings allowing for a controlled ablation down to the submucosa, and the ability of confocal endomicroscopy to provide damage information. Materials and methods: A Nd:YAG laser was applied onto the gastric mucosa of 21 Wistar rats on 10 spots (total 210). Power settings ranging from 0.5 to 2.5W were applied during 1–12 s, with a consequent energy delivery varying from 0.5 to 30 J. Out of the 210 samples, a total of 1050 hematoxilin–eosin stained slides were obtained. To evaluate thermal injury, the ratio between the damage depth (DD) over the mucosa and the submucosa thickness (T) was calculated. Effective and safe ablation was considered for a DD/T ratio ≤ 1 (only mucosal and submucosal damage). Confocal endomicroscopy was performed before and after ablation. A numerical model, using human physical properties, was developed to predict thermal damage. Results: No full-thickness perforations were detected. On histology, the DD/T ratio at 0.5 J was 0.57 ± 0.21, significantly lower when compared to energies ranging from 15 J (a DD/T ratio = 1.2 ± 0.3; p < 0.001) until 30 J (a DD/T ratio = 1.33 ± 0.31; p < 0.001). Safe mucosal and submucosal ablations were achieved applying energy between 4 and 12 J, never impairing the muscularis propria. Confocal endomicroscopy showed a distorted gland architecture. The predicted damage depth demonstrated a significant positive linear correlation with the experimental data (Pearson’s r 0.85; 95% CI 0.66–0.94). Conclusions: Low-power settings achieved effective and safe mucosal and submucosal ablation. The numerical model allowed for an accurate prediction of the ablated layers. Confocal endomicroscopy provided real-time thermal damage visualization. Further studies on larger animal models are required.