Pre-planning Mathematical Models in EUS-guided Laser Ablation of Pancreatic Lesions

Pancreatic ductal adenocarcinoma (PDAC) is the fourth cause of cancer death in Western countries, with an estimation of 57,600 new cases and 47,050 deaths in 2020 in USA. Since the number of deaths due to this lethal disease is rapidly increasing, it is estimated to become the second leading cause of tumor-related death by 2030. The high mortality rate of PDAC is due to several causes, including: i) a poor early diagnosis, ii) the severe biological aggressiveness, and iii) the lack of response to the systemic therapies. At present, surgical resection, i.e., pancreatectomy, represents the only widely accepted treatment option with the potential to increase long-term survival. However, only 20% of the patients are appropriate surgical candidates at the time of the diagnosis. In the majority of the patients, the disease is at an advanced or metastatic stadium, and the surgical resection does not represent a possible option. In these last cases, only palliative solutions are available. Nowadays there is a growing interest in the investigation of novel and alternative therapeutic strategies which could be used in synergy with radiotherapy and chemotherapy. These methods include the inoculation o antitumoral agents within the lesion, through laparoscopic, percutaneous or echoendoscopic (EUS) guidance and locoregional ablation to reduce the tumoral mass. This solution may be particularly beneficial for those patients who are stable after the chemotherapy, without any signs of disease regression neither progression. The ablation can be obtained with different kinds of energy sources, including radiofrequency, microwave and laser. The first study of EUS-guided LA on human patients was conducted by Di Matteo and collaborators who recently performed a prospective study on the feasibility of EUS-guided LA in cases of inoperable and locally advanced PDAC. A total of 9 patients with stage III pancreatic cancer were screened and underwent EUS-guided LA. The fiber was placed in the upper part of the lesion and the following laser power parameters were applied: 2 W for 800 J, 1000 J, or 1200 J; 3 W for 800 J, 1000 J, or 1200 J; 4 W for 800 J, 1000 J, or 1200 J. The application time automatically ranges between 200 and 600 s, according to the power values. No difficulty was experienced in the technical execution of the procedure. The fiber was clearly visible within the targeted lesion, for the entire duration of the application of laser energy. During the procedure, the tip of the fiber was progressively surrounded by a hyperechoic area. This effect did not compromise the endosonographic visualization of the needle inside the lesion. At the end of the ablation, the EUS highlighted a hyperechoic area along the path of the probe, surrounded by non-homogeneous tissue with hyperechoic areas. No relevant complications were found. Control CT scans after 24 hours, 7 days and 30 days showed well-defined coagulative necrotic areas. The average ablation volumes obtained at 800 J decreased by 85% (1.18 cm3 after 24 hours against 0.17 cm3 after 30 days). The average ablation volumes obtained at 1000 J decreased by 76% (3.16 cm3 after 24 hours against 0.80 cm3 after 30 days). Finally, the mean ablation volumes at 1200 J decreased by 44% after 30 days from treatment (1.77 cm3 after 24 hours against 0.98 cm3 after 30 days). According to these results, the 4 W / 1000 J power configuration achieved the largest ablation volume, without clinical complications. In order to perform a study aimed at the complete treatment of the lesion, it is necessary to identify the laser parameters which are specific to the size and location of the lesion. As identified in a recent study by Paiella, the choice of settings and handling of the laser device requires a certain learning curve. Therefore, a predictive mathematical model of the setting parameters can be very beneficial to simplify the ablative procedure. Furthermore, in the absence of correct parameters, several potential complications can occur, such as acute pancreatitis, bleeding from adjacent vascular structures, duodenal perforation or lesion of the biliary tract. The effectiveness of predictive models for LA has already been demonstrated in some ex vivo and preclinical studies, both on the pancreas and on other tissues. Considering all the aspects above, the present protocol presents a prospective interventional study aimed at the analysis and applicability of predictive mathematical models for the calculation of laser settings in the ablation of pancreatic lesion by means of a EUS-guided LA. The investigators will use a mathematical model for the calculation of the laser parameters (power, treatment time), taking into account the size of the tumor and some anatomical characteristics, such as the proximity of blood vessels. The aims of this study fall within the objectives of the European project LASER OPTIMAL (Laser Ablation: Selectivity and Monitoring for Optimal Tumor Removal, GA 759159, http://www.laseroptimal.polimi.it/) of which Professor Saccomandi is Principal Investigator and Dr. Di Matteo is an expert for clinical procedures. This is a prospective interventional study involving two Italian centers, the Operative Digestive Endoscopy UOC of Fondazione Policlinico Universitario Campus Bio-Medico (the "Coordinating Center") and the Department of Mechanical Engineering of Politecnico di Milano. This is a non-controlled, non-randomized interventional feasibility study with a non-commercial purpose. The Operative Digestive Endoscopy UOC of Fondazione Policlinico Universitario Campus Bio-Medico will be responsible for enrolling patients and performing the medical procedures. The Department of Mechanical Engineering of Politecnico di Milano will work on the laser parameters estimation using a mathematical model. Pre-operative radiological images of the candidates will be used for the purpose. Patients with stage III and IV PDAC; non-functioning or functioning neuroendocrine tumors; pancreatic metastases from renal cell carcinoma will be evaluated and included in the study if eligible. All patients will be followed at the Coordinating Center from enrollment to death or the end of the study. Description of study procedures: 1. Preoperative radiological image acquisition. MR images collected within 7 days before the LA EUS-guided procedure will be used by the study collaborators to reconstruct the patient's model. The tumor size and some anatomical features, such as the proximity of blood vessels, will be used in the patient's model. Techniques of image segmentation, 3D geometry reconstruction, and mathematical simulation of laser-tissue interaction will be used to estimate the optimal parameters for the laser (power, energy, time), according to the procedure described by Scott et al. Specifically, Mimics Innovation Suite software will be used to obtain the 3D geometry of the anatomical part of interest (pancreas, tumor, surrounding structures). This will then be imported into finite element computational software (Comsol Multiphysics) allowing the correct and patient-specific settings to be calculated. 2. Estimation of laser parameter settings with the mathematical model The mathematical model is based on finite element simulations implemented using the commercial Comsol Multiphysics software, as previously proposed by Scott et al. Laser settings will be based on values calculated in the Di Matteo and coworkers's experimentation, as previously performed by Saccomandi et al. on ex vivo models. This part of the study will be carried out in collaboration with the Department of Mechanical Engineering of Politecnico di Milano as part of the European LASEROPTIMAL project (www.laseroptimal.polimi.it), coordinated by Paola Saccomandi, and for which Dr. Di Matteo is an external expert. 3. LA EUS-guided EUS will be performed with a FUJIFILM EG- 580UT linear endoscope. Procedures will be performed on the patient under deep sedation, positioned on the left flank. Prophylactic antibiotics will be administered immediately before the procedure and for the following 3 days (ceftriaxone 1 g twice daily intravenously). LA will be performed with a 980 nm laser light (EUFOTON Laser Spectrum) guided by a 300 μm optical fiber (EUFOTON). An 22 Gauge needle (Boston Scientific Expect Slimline) will be used for the optical fiber placement. The protrusion of the fiber end from the needle tip should be 5 mm. After the correct placement of the fiber inside the tumor, the ablation treatment can start. Once inside the lesion, the needle will be slightly retracted, and the fiber will be gently pushed out of the needle tip by a length of 5 mm. The fiber will be placed at the top of the lesion, and the laser will be turned on. The laser settings estimated by the mathematical model will be used. The total number of ablations will depend on the radiological characteristics of the lesion and location. Special attention must be paid to have 1 cm between the fiber tip, the walls of the main vessels, and any metal stents that may be present.