Utility of a prognostic assessment tool to predict survival following surgery for brain metastases

• Published in: Neuro-oncology practice Vol. 10; no. 6; pp. 586 - 591
• Main Authors: Cuthbert, Hadleigh, Riley, Max, Bhatt, Shreya, Au-Yeung, Claudia Kate, Arshad, Ayesha, Eladawi, Sondos, Zisakis, Athanasios, Tsermoulas, Georgios, Watts, Colin, Wykes, Victoria
• Format: Journal Article
• Language: English
• Published: US Oxford University Press 01.12.2023
• Subjects: graded prognostic assessment; prognosis; resection; brain metastases; outcome
• ISSN: 2054-2577; 2054-2585
• DOI: 10.1093/nop/npad047

Abstract Background Brain metastases account for more than 50% of all intracranial tumors and are associated with poor outcomes. Treatment decisions in this highly heterogenous cohort remain controversial due to the myriad of treatment options available, and there is no clearly defined standard of care. The prognosis in brain metastasis patients varies widely with tumor type, extracranial disease burden and patient performance status. Decision-making regarding treatment is, therefore, tailored to each patient and their disease. Methods This is a retrospective cohort study assessing survival outcomes following surgery for brain metastases over a 50-month period (April 1, 2014–June 30, 2018). We compared predicted survival using the diagnosis-specific Graded Prognostic Assessment (ds-GPA) with actual survival. Results A total of 186 patients were included in our cohort. Regression analysis demonstrated no significant correlation between actual and predicted outcome. The most common reason for exclusion was insufficient information being available to the neuro-oncology multidisciplinary team (MDT) meeting to allow GPA calculation. Conclusions In this study, we demonstrate that “predicted survival” using the ds-GPA does not correlate with “actual survival” in our operated patient cohort. We also identify a shortcoming in the amount of information available at MDT in order to implement the GPA appropriately. Patient selection for aggressive therapies is crucial, and this study emphasizes the need for treatment decisions to be individualized based on patient and cancer clinical characteristics.