Prognostic Indicators for Salvage Surgery of Recurrent Sinonasal Malignancy

Daniel James Kaplan,Jee Hong Kim,Eric Wang,Carl Snyderman
DOI: https://doi.org/10.1177/0194599815606699
Abstract:Objective: Identify prognostic factors after salvage surgery for recurrent sinonasal malignancy (SNM). Study design: Case series with chart review. Setting: University of Pittsburgh Medical Center. Subjects: Forty-two patients who underwent curative surgery for locally recurrent SNM ± adjuvant therapy from June 5, 2000, to December 19, 2012. Patients without follow-up were excluded. Methods: Chart review with established prognostic indicators for primary malignancies. Statistical analysis included Kaplan-Meier log-rank test, Fisher's exact test, Student's t test, and Cox regression. Results: Forty-two patients met inclusion criteria: 38.5% developed a second recurrence, and 21.4% had metastases following treatment. The average disease-free interval (DFI) was 26.9 months (range, 2-90 months). DFI was significantly affected by ethmoid versus nonethmoid site (P = .049), histology (P = .012), carotid artery involvement (P = .008), perineural extension (P = .006), and clival invasion (P = .015). The overall survival rates at 6 months, 12 months, and 5 years following surgery were 83.3%, 69%, and 47.6%, respectively. Survival was affected by histology (P = .014), stratified grade (P = .042), tumor extension into the orbit (P = .019), carotid artery (P = .001), perineural space (P = .028), and clivus (P = .022). Complications occurred in 28.6% of patients and were associated with histology (P = .04). Length of hospital stay related to treatment was affected by histology (P = .009), grade (P = .013), and postoperative complication (P < .001). The median percentage of time hospitalized was 8%, and 43% of patients who died within 12 months spent >10% of their remaining days in the hospital. Conclusion: High-risk histologic subtype (melanoma, sinonasal undifferentiated carcinoma, adenocarcinoma, neuroendocrine cancer, sarcoma, and squamous cell carcinoma), grade, and orbital and skull base involvement negatively affect survival and/or DFI for patients with local recurrence of SNM. Improved stratification of patients can be used to guide decision making for patients with recurrent SNM and to avoid inappropriate surgery.
What problem does this paper attempt to address?