Older age is a risk factor for chemotherapy related toxicity. Hurria et al1 published an assessment tool predicting chemotherapy toxicity better than performance status (PS) alone validated in a United States population. A prospective audit to pilot the use of this tool at Wellington Hospital was conducted.

Method:
Adults 65 years with solid tumours commencing a new line of chemotherapy between October 2012 and April 2013 were included. Patient characteristics along with data to allow calculation of Hurria’s predictive tool were recorded, including haemoglobin, creatinine clearance, and 5 geriatric functionality assessment questions. Clinicians were further asked to predict the risk of Grade 3 to 5 toxicity for each patient. Medical documentation was reviewed for any Grade 3 to 5 toxicities. A feasibility questionnaire was also completed by each clinician.

Results:
Of those who have completed treatment (50/65), 4 discontinued early due to clearly progressive disease. 46 (21F, 25M) patients are thus included to date. Median age 73 (range 65 to 90); gastrointestinal cancer (GI) was the most common subtype (43%),.

32 of the 46 patients (70%) experienced grade 3-5 toxicity, compared with 53% in the Hurria study. Preliminary results have not shown a linear relationship between the risk score and toxicity. This may be due to the increased prevalence of GI malignancy in our population (43 v 27%). KPS did not predict toxicity.

Clinicians consistently underestimated potential for grade 3-5 toxicity. Registrars and consultants performed equally poorly. All 11 clinicians found the audit tool easy to use, for 75% of clinicians tool took 5 minutes to complete.

Conclusion:
This pilot confirmed the need for a geriatric chemotherapy toxicity tool. Although the Hurria risk model was easy to use, the estimated score did not correlate with risk of toxicity. Prior to clinical use in NZ, larger studies will be required.