Giuseppe Quero, MD1, Paola Saccomandi, PhD1, Bernard Dallemagne, MD2, Jung-Myun Kwak, MD1, Francesco Maria Di Matteo, MD3, Guido Costamagna, MD1, Jacques Marescaux, MD, FACS, HonFRCS, HonFJSES2, Michele Diana, MD2. 1IHU-Strasbourg, Institute of Image-Guided Surgery, Strasbourg, France, 2IRCAD, Research Institute against Cancer of the Digestive System, Strasbourg, France, 3Campus Bio-Medico di Roma, Roma, Italy
Background: Endoscopic submucosal dissection enables en-bloc removal of early gastrointestinal neoplasms. However, it is technically demanding and time-consuming. Laser-based ablation (LA) techniques, are limited by the lack of depth penetration control and thermal damage (TD) prediction. Our aim was to evaluate a predictive numerical modelling (PNM) of the TD to preoperatively select the optimal power and exposure time enabling a controlled ablation down to the submucosa (SM). Additionally, the ability of confocal endomicroscopy (CE) to provide information on the TD was assessed.
Methods: The stomach of 21 Wistar rats (weight 522±47g) was opened longitudinally. A Nd:YAG continuous laser source (wavelength 1064nm, fiber surface 0.28mm2) was directly applied on the gastric mucosa (M) at 10 randomly chosen spots (total n=210). The energy applied (J=W*s) was randomly set, before each new LA, ranging from values of 2.5-2-1.5-1-0.5W, applied during 12-10-8-6-4-2-1s. This window was pre-determined experimentally, after observing that 3W*1s and 2.5W*13s were constantly leading to a full-thickness perforation. A total of 1050 Hematoxilin-Eosin stained slides, each containing 5 specimens (total n=5250) were assessed to measure the damage ratio (R), defined as the ablation depth over the M+SM tissue thickness, using an image analyzer. Effective and safe ablation (R≤1) was considered when the TD was containing M and SM but sparing muscular and serosa. CE was performed before and after LA. A PNM using finite element analysis with human tissue mechanical and optical properties, was applied and tested for accuracy in predicting the TD.
Results: There were no full-thickness perforations. At the histology, there was an increased damage depth per higher J applications. The R value at 0.5J was 0.57±0.21, and was significantly lower when compared to energies from 15J (R=1.2±0.3; p<0.001) up to 30J (1.33±0.31; p<0.001). Safe M and SM ablations were achieved applying lower P settings (0.5 and 1W), at different t values, leading to an MP impairment only in 5 and 20% of the cases, respectively. CE provided relevant images of the TD, consisting in architecture’s distortion and disappearance of the gland’s contours. The predicted damage depth, demonstrated a significant positive linear correlation with the experimental data (Pearson’s r 0.85; 95% CI 0.66 to 0.94).
Conclusions: Low-power laser settings achieved effective and safe ablation of the M and SM in this experimental model. The PNM enabled an accurate prediction of the future ablated layer and CE enabled real-time visualization of the TD. Further studies on larger animal models are required.
Presented at the SAGES 2017 Annual Meeting in Houston, TX.
Abstract ID: 87650
Program Number: P686
Presentation Session: iPoster Session (Non CME)
Presentation Type: Poster