Better methods needed for predicting poor chemotherapy outcomes in elderly cancer patients

10 Jan 2021
In light of its ageing population, Singapore seeks to develop more innovative means to cope with rehab and nursing.In light of its ageing population, Singapore seeks to develop more innovative means to cope with rehab and nursing.

Neither the Comprehensive Geriatric Assessment (CGA) tool nor an oncologist’s clinical judgement can reliably predict unfavourable outcomes in a heterogenous population of elderly cancer patients, a recent study reports.

Moreover, the CGA and clinical judgement are also not well-aligned, yielding discordant assessments in more than a third of cases.

The researchers conducted a prospective cohort study on 110 consecutive elderly cancer patients. All participants were assessed in accordance with the CGA, while 66 patients received their medical oncologists’ estimation of their tolerance of the standard treatment. Unfavourable outcomes were defined as toxicities of at least grade 3, dose reductions, treatment postponements, early progression, and death before treatment initiation.

Ultimately, the analysis of unfavourable outcomes could not be carried out in 30 patients because they did not undergo anticancer therapy. In total 80 patients were assessed for overall unfavourable outcomes, while 60 were evaluated for chemotherapy-related outcomes.

Sixty percent (n=48) of the overall sample and 75.0 percent (n=45) of those treated with chemotherapy developed an unfavourable outcome. CGA alone proved to be underpowered for predicting unfavourable outcomes, with an area under the curve (AUC) of 0.51; its sensitivity and specificity values were likewise middling or low, at 64.6 percent and 37.5 percent, respectively.

Similarly, the oncologists’ clinical assessment could not reliably predict the occurrence of unfavourable outcomes, yielding an AUC of 0.57, a sensitivity of 40.7 percent, and a specificity of 72.2 percent. The combination of both CGA and professional judgement was likewise insufficient (AUC, 0.51; sensitivity: 29.6 percent; specificity: 72.2 percent).

“The results emphasize the need for more objective measures to identify older patients with a high risk of unfavourable outcome to provide reasoning for refraining from or adapting treatment,” the researchers said.

J Geriatr Oncol 2021;12:49-56