Assessment of viability of human fat injection into nude mice with micro-computed tomography.

Lipotransfer is a vital tool in the surgeon's armamentarium for the treatment of soft tissue deficits of throughout the body. Fat is the ideal soft tissue filler as it is readily available, easily obtained, inexpensive, and inherently biocompatible.(1) However, despite its burgeoning popularity, fat grafting is hampered by unpredictable results and variable graft survival, with published retention rates ranging anywhere from 10-80%. (1-3) To facilitate investigations on fat grafting, we have therefore developed an animal model that allows for real-time analysis of injected fat volume retention. Briefly, a small cut is made in the scalp of a CD-1 nude mouse and 200-400 µl of processed lipoaspirate is placed over the skull. The scalp is chosen as the recipient site because of its absence of native subcutaneous fat, and because of the excellent background contrast provided by the calvarium, which aids in the analysis process. Micro-computed tomography (micro-CT) is used to scan the graft at baseline and every two weeks thereafter. The CT images are reconstructed, and an imaging software is used to quantify graft volumes. Traditionally, techniques to assess fat graft volume have necessitated euthanizing the study animal to provide just a single assessment of graft weight and volume by physical measurement ex vivo. Biochemical and histological comparisons have likewise required the study animal to be euthanized. This described imaging technique offers the advantage of visualizing and objectively quantifying volume at multiple time points after initial grafting without having to sacrifice the study animal. The technique is limited by the size of the graft able to be injected as larger grafts risk skin and fat necrosis. This method has utility for all studies evaluating fat graft viability and volume retention. It is particularly well-suited to providing a visual representation of fat grafts and following changes in volume over time.