LAUSR

BACKGROUND A geriatric assessment (GA) is increasingly used to help guide treatment decisions in older patients with cancer. However, there is no consensus regarding which domains should be included in the GA. In addition, the field of geriatric oncology moves very fast and as a result many new studies have been published since the last review in 2015. Therefore, the objective of this systematic review is to evaluate which domains of the GA could predict patient-related treatment outcomes of older patients with cancer and thereby should be included in a GA. METHODS A systematic literature search was performed for publications in English or Dutch between September 2006 and July 2017 addressing the association between individual domains of the GA and mortality, postoperative complications, or systemic treatment-related outcomes in older patients with cancer. RESULTS Eight different domains were evaluated in 46 publications, namely functional status, nutritional status, cognition, mood, physical function, fatigue, social support, and falls. All eight domains were predictive for at least one of the investigated outcomes but the results were quite variable across studies. Physical function and nutritional status were the domains most often associated with mortality and systemic treatment-related outcomes, and the domain physical function was most often associated with postoperative complications. CONCLUSION Overall, this review demonstrates that the GA should minimally consist of physical function and nutritional status, when the aim is to predict patients-related outcomes of older patients with cancer, although the results are quite heterogeneous. For the other domains, the findings are too inconsistent to draw conclusions about their overall predictive ability.