NIR Device for Assessment of Skin Flap Perfusion in Breast Surgery: A Pilot Interventional Study (ADisIRast)

The ADISIRAST study is a single-center, non-randomized, pre-market, non-profit clinical investigation of a non-invasive near-infrared spectroscopy (NIRS) imaging device for the assessment of perfusion in mastectomy skin flaps in patients undergoing breast surgery with immediate reconstruction.

Mastectomy skin-flap viability is an important determinant of postoperative healing and reconstructive outcomes. In routine clinical practice, flap perfusion is mainly assessed through intraoperative inspection and surgical judgment. Objective methods for evaluating tissue oxygenation may provide complementary information and may help characterize regional perfusion patterns that are not readily apparent during standard clinical assessment.

Near-infrared spectroscopy is a non-invasive optical technique that estimates superficial tissue oxygen saturation (StO₂) by analyzing the relative absorption of near-infrared light by oxygenated and deoxygenated hemoglobin. The investigational device is a portable, non-contact NIRS imaging system that generates two-dimensional maps of superficial tissue oxygen saturation. The device does not require tissue contact, contrast-agent injection, or exposure to ionizing radiation. Acquisitions are performed by positioning the device at a standardized distance from the surgical area and obtaining a digital image together with the corresponding StO₂ map.

The study was designed to evaluate the feasibility, repeatability, and clinical interpretability of NIRS-derived StO₂ measurements in breast skin flaps throughout the perioperative period. NIRS images were acquired at predefined study time points, including a preoperative baseline assessment, an intraoperative assessment, and postoperative follow-up assessments(1 day and 15 days postoperative). This approach allowed the characterization of temporal changes in tissue oxygenation and the evaluation of regional differences within the skin flap.

Image analysis included assessment of absolute StO₂ values, regional oxygenation differences, spatial distribution of tissue areas within predefined StO₂ ranges, and longitudinal changes in oxygenation patterns. Regions of interest were defined within superior and inferior portions of the breast skin flap to explore regional perfusion heterogeneity. Additional image-derived metrics were calculated to quantify the extent of tissue area within selected StO₂ ranges and the degree of spatial heterogeneity of oxygenation.

Clinical and surgical data were prospectively collected, including reconstructive information, perioperative characteristics, postoperative flap healing, and clinically observed complications. NIRS findings were subsequently evaluated in relation to postoperative clinical observations, particularly the presence, location, and evolution of areas showing impaired healing or suspected ischemic changes.

Surgical planning and intraoperative decisions, including the choice of reconstructive approach, were made according to standard clinical practice, based on the treating surgical team's clinical assessment and preoperative evaluation. The operating surgeon was blinded to the intraoperative NIRS images, and NIRS findings did not influence surgical decision-making or the selection of the reconstructive procedure.

After study completion, NIRS-derived perfusion findings were retrospectively compared with the reconstructive strategy selected during surgery. This exploratory analysis was performed to assess the concordance between NIRS-based indications of flap perfusion and the clinical decision made by the surgical team. The analysis was intended to explore the potential future role of non-invasive tissue oxygenation imaging as an adjunctive tool for perioperative assessment and surgical training, rather than to evaluate a device-guided management strategy.

The study does not compare alternative surgical treatments or reconstructive techniques. It is designed as a feasibility and performance investigation and is not powered to establish definitive diagnostic thresholds, predict clinical outcomes with definitive accuracy, or demonstrate superiority of one reconstructive strategy over another. Data were collected prospectively and managed using coded participant identifiers.