Use of CARG toxicity tool to predict chemotherapy toxicity in older veterans to facilitate better treatment outcomes.

Abstract

244 Background: Traditionally the Eastern Cooperative Oncology Group (ECOG) performance status has been used to assess objective patient well- being and to determine whether geriatric patients can tolerate chemotherapy. Hurria and colleagues from The Cancer and Aging Research Group (CARG) developed a validated geriatric assessment model for adults > 65 yrs. to predict the risk of grade 3-5 chemotherapy toxicity in the treatment of solid tumors. The CARG score, ranging from scores of 0 to 19, incorporates measures of functional status such as fall history, hearing problems, physical capabilities, performance status as well as objective measures including age, gender, height, weight, cancer type, type and dose of chemotherapy, hemoglobin, and creatinine clearance. Methods: At the Dayton VA Medical Center, the utility of the CARG tool was studied in patients to predict chemotherapy tolerance for treatment naïve veterans age > 65 yrs. who were diagnosed with solid tumor malignancies excluding leukemias. CARG scores were stratified as low (score 0-5), intermediate (score 6-9) and high risk (score 10-19). The CARG toxicity score was compared with standard use of ECOG performance status alone to predict chemotherapy toxicity. Subsequent changes in chemotherapy dosage, treatment delays due to adverse effects as well as mortality rate after treatment were also monitored. Results: The study, which enrolled 20 patients over 1 year, revealed that out of 18 patients with ECOG 0-2, 94% had at least an intermediate CARG risk for chemotherapy toxicity and 50% were shown to have high CARG risk for chemotherapy toxicity. Of the 11 high-risk CARG patients in the study, 45% had chemotherapy intolerance. Of the 17 treated patients with calculated CARG toxicity risk 50% or higher, there was a 41% mortality rate. Conclusions: Geriatric assessment modalities are integral to predict chemotherapy toxicity in geriatric patients. These modalities should be part of routine assessment of geriatric cancer patients in order to adequately incorporate their functional status when calculating chemotherapy dosage as the solitary use of ECOG performance status appears to be inadequate for this purpose. The calculated CARG risk percentage may be a better predictor of mortality rate than the ECOG score.