Characteristics of Patients Who Survived < 3 Months or > 2 Years after Surgery for Spinal Metastases: Can We Avoid Inappropriate Patient Selection?
Jorrit-Jan Verlaan,David Choi,Anne Versteeg,Todd Albert,Mark Arts,Laurent Balabaud,Cody Bunger,Jacob Maciej Buchowski,Chung Kee Chung,Maarten Hubert Coppes,Hugh Alan Crockard,Bart Depreitere,Michael George Fehlings,James Harrop,Norio Kawahara,Eun Sang Kim,Chong-Suh Lee,Yee Leung,Zhongjun Liu,Antonio Martin-Benlloch,Eric Maurice Massicotte,Christian Mazel,Bernhard Meyer,Wilco Peul,Nasir A. Quraishi,Yasuaki Tokuhashi,Katsuro Tomita,Christian Ulbricht,Michael Wang,F. Cumhur Oner
DOI: https://doi.org/10.1200/jco.2015.65.1497
IF: 45.3
2016-01-01
Journal of Clinical Oncology
Abstract:PURPOSE:Survival after metastatic cancer has improved at the cost of increased presentation with metastatic spinal disease. For patients with pathologic spinal fractures and/or spinal cord compression, surgical intervention may relieve pain and improve quality of life. Surgery is generally considered to be inappropriate if anticipated survival is < 3 months. The aim of this international multicenter study was to analyze data from patients who died within 3 months or 2 years after surgery, to identify preoperative factors associated with poor or good survival, and to avoid inappropriate selection of patients for surgery in the future.PATIENTS AND METHODS:A total of 1,266 patients underwent surgery for impending pathologic fractures and/or neurologic deficits and were prospectively observed. Data collected included tumor characteristics, preoperative fitness (American Society of Anesthesiologists advisory [ASA]), neurologic status (Frankel scale), performance (Karnofsky performance score [KPS]), and quality of life (EuroQol five-dimensions questionnaire [EQ-5D]). Outcomes were survival at 3 months and 2 years postsurgery. Univariable and multivariable logistic regression analyses were used to find preoperative factors associated with short-term and long-term survival.RESULTS:In univariable analysis, age, emergency surgery, KPS, EQ-5D, ASA, Frankel, and Tokuhashi/Tomita scores were significantly associated with short survival. In multivariable analysis, KPS and age were significantly associated with short survival (odds ratio [OR], 1.36; 95% CI, 1.15 to 1.62; and OR, 1.14; 95% CI, 1.02 to 1.27, respectively). Associated with longer survival in univariable analysis were age, number of levels included in surgery, KPS, EQ-5D, Frankel, and Tokuhashi/Tomita scores. In multivariable analysis, the number of levels included in surgery (OR, 1.21; 95% CI, 1.06 to 1.38) and primary tumor type were significantly associated with longer survival.CONCLUSION:Poor performance status at presentation is the strongest indicator of poor short-term survival, whereas low disease load and favorable tumor histology are associated with longer-term survival.