Regenerated Oxidized Cellulose (Surgicel) Induces Nasal Epithelial Necrosis In Vivo by Acidifying the Cellular Environment

Introduction: Despite major progress in closure techniques for endoscopic endonasal skull base surgery (EESBS), cerebrospinal fluid (CSF) leakage remains a common complication. Regenerated oxidized cellulose (ROC) (Surgicel¸ Ethicon, United States) sheets are often used as an adjunct to reconstruction to provide hemostasis and promote adherence of reconstructive tissues. However, evidence supporting its impact on the healing process is uncertain and several publications have raised concerns about its possible cellular toxicity. Because it is an organic acid, we hypothesized that ROC can induce cellular necrosis by lowering the pH of the adjacent tissues.

Methods: In five patients, a 1-cm² piece of ROC gauze was placed on the middle turbinate or sphenoid mucosa 15 minutes before they were resected. After removal, mucosa treated with ROC was separated from untreated mucosa (control) and the samples were randomly assigned an identification number to avoid bias during laboratory testing. The investigators processed, analyzed and interpreted the samples blindly. A histological examination of structural changes in the respiratory epithelium including the loss of normal cilia, detachment from the basal membrane and evident cellular necrosis was performed. To assess the effect of ROC on pH, increasing amounts of ROC was added to Dulbecco’s modified medium (DMEM) and a dose–response curve was constructed.

Results: Compared with unexposed controls, treated tissue exhibited a higher incidence of epithelial necrosis (0 vs. 100%, p < 0.05) and epithelial cell detachment from the basal membrane (0 vs. 80%, p < 0.05). Intact epithelium was present in all control samples but only in 20% of samples treated with ROC (p < 0.05). Although loss of cilia was more common in exposed samples (20 vs. 80%), the difference was not significant (p = 0.2). When added to DMEM, ROC caused a dose-dependent decrease in the pH of the medium. Control samples had a pH of 7.65 ± 0.18, whereas adding 1.5, 3, 4.5, 6 and 7.5 mg/mL lowered the pH to 7.1 ± 0.28, 6.56 ± 0.2, 6.29 ± 0.1, 5.78 ± 0.47 and 5.01 ± 0.85, respectively (p < 0.05). Cultured fibroblasts exposed to DMED pretreated with ROC exhibited extensive necrosis; but when the same pretreated medium's pH was taken back to physiologic values, it had no effect on cultured cells.

Conclusion: ROC applied in vivo to the nasal mucosa was demonstrated to induce epithelial necrosis and cellular detachment, and was associated with the absence of structurally normal respiratory epithelium. ROC diminished the medium pH in a dose-dependent fashion which is likely the action mechanism observed in vivo. Use this material may actually inhibit rather than aid in effective repair. Further studies, which examine the impact of ROC on wound healing in vivo, are warranted.

Fig. 1 Effect of ROC on nasal mucosa. (A) Control and (B) treated samples were stained with cytokeratin antibodies (green) marking epithelial cells and vimentin antibodies (red) marking fibroblast. Treated samples showed cells with smaller cytoplasm, detachment from the basal membrane and loss of cilia. e, epithelium; s, stroma; c, cilia.

Table 1 Effect of ROC on the nasal mucosa. Results are expressed as a proportion of the samples positive for a given histological change

No conflict of interest has been declared by the author(s).