75. Preliminary results of the prospective observational study of spinal metastasis treatment (POST)

Abstract

BACKGROUND CONTEXT Several retrospective series support the use of operative or nonoperative interventions for the treatment of patients with spinal metastases. While surgery in several contexts has been shown to result in superior outcomes regarding longevity and ambulatory function, some of these conclusions may result from selection and indication bias. While nonoperative strategies avoid the complications of surgery, treatment results may not always be as efficacious and surgical intervention following nonoperative care has been shown to have inferior results. At present, there is no rigorously performed prospective study focused on functional outcomes for patients undergoing operative and nonoperative management for spinal metastases. PURPOSE We present preliminary results of an ongoing prospective observational study contrasting functional outcomes between patients receiving surgery, or nonoperative care, for spinal metastatic disease. STUDY DESIGN/SETTING Prospective observational study conducted at two academic medical centers (2017-present). PATIENT SAMPLE A total of 178 patients receiving operative or nonoperative treatment for spinal metastases. OUTCOME MEASURES Enrolled patients receive an initial intake evaluation, including assessment of demographics, past medical history, surgical history, radiology, primary tumor histology, ambulatory status and functional evaluation. The functional evaluation consists of study-specific surveys: EuroQuol 5-Dimension (EQ5D) profile and Short-Form (SF)-12 (assessment of state of health and physical/mental function), visual analog scale (VAS) for pain (assessment of current level of pain) and PROMIS Measures (assessment of global health). Completion of the EQ5D, SF-12, VAS pain scale, PROMIS (appendix) and a determination of ambulatory status will occur at the 1-month, 3-month, 6-month and 12-month time-points. METHODS Survival is the principle dependent variable with clinically meaningful survival defined as three months or greater. Treatment-associated complications and functional outcomes at time-points up to one year will be considered secondary dependent variables. Patients will be categorized into groups based on the clinically meaningful time-points of 3 months survival, 6 months survival or survival to one year or greater. Patients who received surgery will be compared to those managed nonoperatively using propensity-matched techniques. The propensity score will be developed for surgical intervention as the dependent variable using a logistic regression model. Patient matching will be made based on a propensity score. The capacity of the predictive score to inform outcomes in the setting of surgical or nonoperative intervention will then be evaluated in the propensity matched model and also using a mixed-effects model. RESULTS Phase I enrollment consists of 178 patients. The average age of the cohort is 60.4 (SD 13.1). Fifty-six percent of the population is male. Eighty-five percent of the population is white. Average body mass index is 27.0 and average co-morbidities are 2.4. Breast (19%) and lung cancer (18%) are the most common primary tumors. Thirty-seven percent of metastases were isolated to the thoracic region and 17% in the lumbar region, while 33% involved multiple regions. Thirty-seven percent of the cohort has been treated surgically. At present, 58% of the population is still alive. Surgical intervention (24/65; 37% mortality) is not currently associated with superior survival as compared to nonoperative treatment (50/109; 46% mortality; p=0.25). CONCLUSIONS We anticipate that the results of this analysis will provide high quality evidence that accurately identifies those patients who will maximally benefit from surgical intervention with the lowest risk of perioperative morbidity. FDA DEVICE/DRUG STATUS This abstract does not discuss or include any applicable devices or drugs.