Poster Presentation Clinical Oncology Society of Australia Annual Scientific Meeting 2017

Predicting chemotherapy toxicity in older adults: comparing the value of the CARG Toxicity Score with oncologists’ estimates of toxicity (#189)

Erin B Moth 1 , Belinda E Kiely 1 2 , Vasikaran Naganathan 3 , Andrew Martin 2 , Natalie Stefanic 4 , Philip Beale 1 , Peter Grimison 4 5 , Prunella Blinman 1
  1. Concord Cancer Centre, Concord, NSW, Australia
  2. NHMRC Clinical Trials Centre, University of Sydney, Sydney, NSW, Australia
  3. Centre for Education and Research on Ageing, Concord Repatriation General Hospital, Concord, NSW, Australia
  4. University of Sydney, Sydney, NSW, Australia
  5. Chris O'Brien Lifehouse, Sydney, NSW, Australia


The CARG Toxicity Score predicts severe (grade ≥3) chemotherapy-related toxicity in adults ≥65yrs by classifying patients as low (score 0-5), intermediate (score 6-9), or high-risk (score 10-23).(1,2) It has not been validated in the Australian setting. We determined the predictive value of the CARG Score, and compared it to oncologists’ clinical estimates for severe chemotherapy toxicity in older adults with cancer.


The CARG Score (0-23) was calculated for patients ≥65yrs starting chemotherapy for a solid organ cancer (any type/stage). Treating oncologists independently estimated the probability of severe chemotherapy toxicity (0-100%). The predictive value of the CARG Score, oncologists’ estimates, and the two measures combined were determined by logistic regression and area under the receiver operating characteristic (AU-ROC) curves.


126 patients from 10 oncologists over 2 sites participated. Median age was 72yrs (range 65-84). Most patients were male (75, 60%), having palliative chemotherapy (82, 65%), most commonly for colorectal cancer (45, 36%). The median CARG Score was 7 (range 0-17), with 25 (20%) classified as low, 77 (61%) intermediate, and 24 (19%) high-risk. The median oncologist estimate of severe chemotherapy toxicity was 30% (range 3-80%). 64 patients (52%) experienced a grade ≥3 toxicity over a median of 6 cycles (range 1-17). The CARG Score did not predict severe toxicity (OR 1.04, 95%CI 0.92-1.18, p-value 0.54, AU-ROC 0.52); rates of severe toxicity in low, intermediate, and high-risk groups (by CARG Score) were 58%, 47%, and 58% respectively. Oncologists’ estimates did not predict severe toxicity (OR 1.00, 95%CI 0.98-1.02, p-value 0.82, AU-ROC 0.52), nor did a model combining the two measures (AUC-ROC 0.52).  


The CARG Score, oncologists’ estimates, or a combined measure of the two did not predict severe chemotherapy-related toxicity in our local population of older adults. Methods to improve risk prediction in the local setting are needed.

  1. 1. Hurria A, Mohile S, Gajra A, Klepin H, Muss H, Chapman A, et al. Validation of a Prediction Tool for Chemotherapy Toxicity in Older Adults With Cancer. J Clin Oncol. 2016.
  2. 2. Hurria A, Togawa K, Mohile SG, Owusu C, Klepin HD, Gross CP, et al. Predicting chemotherapy toxicity in older adults with cancer: a prospective multicenter study. J Clin Oncol. 2011;29(25):3457-65.