An attempt to correlate a "Multidimensional Geriatric Assessment" (MGA), treatment assignment and clinical outcome in elderly cancer patients: Results of a phase II open study

Objectives: To establish a correlation between a specific MGA category, an appropriate preventively established treatment and clinical outcome in a population of elderly cancer patients. The ultimate goal was to verify whether the appropriate treatment given to elderly cancer patients according to their MGA category could translate into a better clinical outcome assessed as clinical response and toxicity, i.e whether this process might achieve a clinically meaningful impact. Patients and methods: We carried out a phase II open, prospective non-randomized study in 75 elderly cancer patients (lung, head and neck, colorectal, gynecologic and breast) hospitalized at the Department of Medical Oncology, University of Cagliari, Italy. All patients underwent MGA evaluation and were assigned to three different categories: fit, intermediate and frail. Thereafter, an appropriate preventively established treatment was administered and the clinical outcome was assessed. The clinical outcome after 3 month treatment was defined on the basis of objective clinical response and toxicity. The difference of clinical outcome in the MGA categories was assessed by ANOVA test. Moreover, the correlation between MGA category and the clinical outcome (clinical response and toxicity) was assessed by Spearman's correlation test. Results: A better clinical response was observed in fit patients as compared both to intermediate and frail patients. Treatment toxicity was comparable in the different MGA categories. The correlation analysis between MGA category, clinical response to treatment and toxicity showed that there was a significant direct correlation with clinical response and no correlation with toxicity. Overall, the regression analysis showed that MGA was predictive of clinical outcome, in the sense that it is truly predictive for clinical response and no predictive for toxicity. Conclusion: Our study demonstrates that the MGA, although time-consuming, is a useful and cost-benefit effective tool to appropriately select elderly cancer patients to be treated effectively in terms of a survival advantage and those who would benefit mainly in terms of improvement of quality of life. Moreover, the treatment preventively established for each MGA category was shown to be adequate and accomplished the most appropriate performances in terms of effectiveness and toxicity. (C) 2007 Elsevier Ireland Ltd. All rights reserved.