Anterior Skull Reconstruction: Does Fat Preparation Matter?

Background: Anterior skull base defects are a part of many surgical approaches to neoplasms, inflammatory diseases and trauma of this part of the anatomy. CSF rhinorrhea may result from trauma, surgery, or rupture of preexisting defect in the skull base. Many materials have been used to reconstruct skull base defects, but the most universal material, whether by itself or in combination with other tissue, is subcutaneous fat. It has not been validated whether the fat needs any special preparation before implantation. Establishing whether unprepared fat is equally effective as specially prepared fat may result in quicker surgical times and the need to harvest a smaller volume of fat, both leading to better patient outcomes. We sought to determine if dried and compressed fat leads to better surgical success rates.

Methods: After approval from our institution's IACUC, we created anterior skull base defects in 28 Sprague Dawley rats based on a previously validated model. Reconstruction was then done in three arms: skin closure only, native fat sealed with fibrin glue, and specially prepared fat sealed with fibrin glue. The preparation of the fat included drying and compressing the native fat. After 2 weeks of healing time, fluorescein was injected into the subarachnoid space through a burr hole and allowed to circulate while the animal was under anesthesia. The animals were then euthanized, and the nasal bones were removed. The persistent CSF leak was determined by the presence of fluoresce on the nasal mucosa.

Results: The control group had a CSF leak rate of 67% (n = 9), the native fat group had a 44% leak rate (n = 9), and those with prepared fat had a 30% leak rate (n = 10). Chi-squared analysis showed that these differences were not statistically significant. The p value between the control and prepared fat groups was p < 0.11, p < 0.35 between the control and native fat groups, and p < 0.52 between the native and prepared fat groups. The surgical time averaged 16 minutes for the control group, 20 minutes for the native fat, and 22 minutes for the prepared fat group. One-way ANOVA demonstrated that the surgical times were significant when comparing the control group to the two reconstruction arms, but the native and prepared fat reconstruction groups were not different from each other.

Discussion: Using dried and compressed fat leads to decreased rates of CSF leaks, but this decrease did not reach statistical significance. Furthermore, the difference in surgical times between the native and prepared fat did not reach significance. These results were a result of small sample size. On the basis of our first set of data, we expect roughly 75% success in the dry group and 30% success in the control groups. On the basis of a confidence interval of 95% and power of 75%, we need a total of 14 animals per group. We are furthering this project by adding these additional animals, and final data will be available for presentation in February 2015.