Predicting chemotherapy toxicity in older adults with lung cancer

Existing oncology performance status measurements are used to predict chemotherapy toxicity in all patients with cancer, regardless of age. A new predictive model for grade 3-5 chemotherapy toxicities was developed by Hurria et al. (2011).(1) As the model is from the Cancer and Aging Research Group (CARG), we call it the CARG toxicity tool. We investigated whether this tool can usefully characterize chemotherapy risks for older patients with lung cancer.

Patients from our hospital aged ≥ 65 years with lung cancer completed a questionnaire form prior to chemotherapy. We reviewed patients' chemotherapy courses to identify toxicities, and used the toxicity tool to score the patients' outcomes. The sample was divided into three risk strata based on approximate risk score quartiles, with the middle two quartiles combined. Chi-square statistics were used to verify differences among groups.

Between September 2011 and September 2012, 120 patients with lung cancer (87 males and 33 females) ≥ 65 years of age (mean: 69 years; range: 65-82 years) were enrolled in the study. In our sample, 35% of subjects had ≥ one grade 3-5 hematologic toxicity; 48% had ≥ one grade 3-5 nonhematologic toxicity. Toxicity varied significantly among the risk groups (P<0.001), but the incidence of toxicity did not vary significantly among the KPS-based risk groups (P=0.322).

This new CARG toxicity tool can be used to better distinguish the risks of chemotherapy toxicity than the KPS for older patients with lung cancer, and may change the standards for oncology assessments.