We evaluated vessel counts in the pharyngeal mucosal margins of patients who underwent salvage laryngectomy to establish whether mucosal vascularity might predict fistula risk. Retrospective cohort. Tertiary Medical Center. Patients who underwent salvage total laryngectomy at our institution between 1999 and 2015 were identified. Pharyngeal mucosal margins from laryngectomy specimens were evaluated histologically for each patient, and vessel counts were performed on 5 ×10 images. The primary outcome measure was fistula within 30 days of surgery and mean vessel counts were assessed as the principle explanatory variable. Seventy patients were included and 40% developed a postoperative fistula. There was a large difference in the mean vessel count in patients who did develop fistula (48.6 vessels/×10 field) compared to those who did not (34.7 vessels/×10 field). A receiver operative characteristic curve found that a cutoff value of 33.9 vessels/×10 field provided a sensitivity of 75% and specificity of 62% to predict the likelihood of fistula occurrence (area under the curve = 0.71, 95% confidence interval [CI]: 0.59-0.83). In a binary logistic regression, patients with vessel counts greater than 33.9 had a 5-fold increased risk of developing fistula (95% CI: 1.8-16.45). Histologically, vessels in the pharyngeal mucosa of patients who developed fistulas were more disorganized. After salvage laryngectomy, patients with higher mean mucosal margin vessel counts are at increased risk of fistula. The mechanism is unknown, but the disorganization of the vasculature may contribute to poor wound healing. Vessel counting may allow for fistula risk stratification and guide postoperative care.
Read full abstract