Nicholas Navetta
Research Mentor: Alvaro Rojas-Peña
Mentor Department: Department of Surgery-Transplantation, Medicine
Author(s): Nicholas Navetta, Louis Beardell III, Vinisha Somaya, Hannah Naldrett, Sikander Raza, Gergely Lautner, Alvaro Rojas-Peña
Session: Session 4 (1:00 PM – 1:50 PM)
Presentation Type: Poster 127
Abstract
High-flow nasal cannula (HFNC) therapy is a well-established, non-invasive method for improving oxygenation in patients with acute hypoxia when standard low-flow nasal cannulas are insufficient. Prior to clinical translation, preclinical?in vivo?evaluation is required; however, human-designed cannulas poorly fit ovine anatomy and lack durability for chronic studies. We demonstrate that a customized nasal cannula enables effective, repeatable oxygen delivery in an ovine HFNC model. Methods: High-flow nasal cannulas were modeled (Fig. 1), characterized using computational fluid dynamics (CFD), and fabricated using stereolithography (SLA) 3D printing with Formlabs BioMed Elastic 50A resin. A Telesair Bonhawa high-flow nasal cannula system was configured to deliver heated (37 °C), humidified (100%) gas at 40 L/min with 80% FiO2. In vivo?testing was conducted in five juvenile sheep (n = 5) using a chronic, sutured configuration to assess oxygen delivery, flow obstruction characteristics, and device durability. Arterial pO2, FiO2, flow obstruction events, qualitative fit, and post-study mucus accumulation were evaluated. Results: CFD analysis was used to optimize cannula design, and benchtop testing demonstrated minimal resistance (< 4 mmHg). Sheep (44.2±7.9 kg) were successfully instrumented for invasive hemodynamic monitoring, and the nasal cannula was secured using sutures and a headband. Total support time across all animals was 72 hours. Three FiO2 levels were tested over the study duration: Period 1 (hours 15–20) at 20% FiO2, Period 2 (hours 20–68) at 80% FiO2, and Period 3 (hours 68–72) at 20% FiO2. For arterial pO2, Period 1 showed a % error of 11.4%, Period 2 showed a % error of 17.7% (including an outage), and Period 3 showed a % error of 4.5%. A temporary loss of building oxygen supply at hour 45 resulted in a transient reduction in both FiO2 and pO2 during Period 2. Overall, all sheep tolerated the nasal cannulas well (e.g., no impairment of feeding or drinking), and post-necropsy evaluation revealed no obstruction, or device failure. Conclusion: Custom 3D-printed nasal cannulas were well tolerated and effectively provided consistent oxygen therapy in an ovine preclinical model. Even after 72 hours of continuous use, the cannula maintained effective oxygen delivery. Considerations such as internal and external mucus buildup, ventilation, comfort, and resistance were critical in creating an effective, low-cost device.
