BACKGROUND AND OBJECTIVES: Whether unintended durotomy in metastatic spine surgery seeds tumor in the central nervous system (CNS) remains unknown. Our objectives were to (1) determine the rate of unintended durotomy, (2) identify the preoperative and perioperative factors that increase the risk of unintended durotomy, and (3) determine whether unintended durotomy affected local recurrence (LR), any spinal recurrence, and overall survival. METHODS: A single-center, retrospective cohort study of patients undergoing metastatic spine surgery between January 2010 to January 2021 was undertaken. The primary exposure variable was the occurrence of unintended durotomy. Multivariable logistic/linear regression were performed controlling for age, body mass index, tumor size, other organ metastases, and preoperative radiotherapy/chemotherapy. RESULTS: Of 354 patients undergoing surgery for extradural spinal metastasis, 19 patients (5.4%) had an unintended durotomy. Preoperatively: No difference was found between patients with and without unintended durotomy regarding basic demographics, comorbidities (P = .645), or tumor histology (P = .642). Preoperative chemotherapy/radiotherapy were similar between the 2 groups. Perioperatively: Although patients with unintended durotomy had more costotransversectomies (36.8% vs 12.8%, P = .010), no difference was found in tumor characteristics, operative time (337.4 vs 310.6 minutes, P = .150), blood loss (1012.8 vs 883.8 mL, P = .157), length of stay (6.4 vs 6.9 days, P = .452), or overall reoperation/readmission. Long-term: No difference was seen in CNS spread between those with unintended durotomy and no durotomy (10.5% vs 3.0%, P = .077). LR (5.3% vs 12.2%, P = .712), time to LR (388.0 vs 213.3 ± 359.8 days, log-rank; P = .709), any spinal recurrence (26.3% vs 34.0%, P = .489), overall survival (21.05% vs 34.3%, P = .233), and time to death (466.9 ± 634.7 vs 465.8 ± 665.4 days, log-rank; P = .394) were similar on multivariable Cox regressions. CONCLUSION: In patients undergoing surgery for extradural spinal metastases, 5% had an unintended durotomy, and costotransversectomies were associated with increased risk of an unintended durotomy. Unintended durotomies did not lead to increased LR or shorter survival. Taken together, shortened survival due to seeding tumor into the CNS after an unintended durotomy was not observed.
Read full abstract