Abstract
PATIENTS AND METHODS
A total of 1463 patients previously included in a prospective cohort study and treated with standard-of-care capecitabine-based anticancer regimens (monotherapy or combined with other chemotherapy or radiotherapy) were analysed. Logistic regression models were developed to investigate associations between patient- and treatment-related factors and occurrence of early--i.e. cycle one or two--severe (grade ≥ 3) treatment-related toxicity, toxicity-related hospitalisation, and toxicity-related treatment discontinuation. Performance of models was evaluated using receiver-operating characteristic (ROC) curves and internal validity was explored using bootstrap analysis.
CONCLUSION
Renal function, BSA, and age, in addition to pre-treatment uracil, are associated with clinically relevant differences in risk of early severe toxicity in patients treated with capecitabine in routine clinical care.
RESULTS
Among 1463 patients included, 231 patients (16%) experienced early severe toxicity, 132 patients (9%) were hospitalised for toxicity, and 146 patients (10%) discontinued treatment for toxicity; in total, 321 patients (22%) experienced any early toxicity-related adverse outcome. Predictors of early grade ≥ 3 toxicity, after adjustment for treatment regimen, were renal function (odds ratio [OR] 0.85 per 10 ml/min/1.73 m(2), p = 0.0007), body surface area (BSA) (OR 0.33 per m(2), p = 0.0053), age (OR 1.14 per decade, p = 0.0891), and elevated pre-treatment uracil concentrations (OR 2.41 per 10 ng/ml, p = 0.0046). Age was significantly associated with fatal treatment-related toxicity (OR 5.75, p = 0.0008). Area under the ROC curve (AUC) of a model to predict early grade ≥ 3 toxicity was 0.704 (95% confidence interval 0.666-0.743, optimism-corrected AUC 0.690).
BACKGROUND
The objective of this analysis was to determine the factors associated with early onset treatment-related toxicity in patients treated with capecitabine-based anticancer regimens in daily clinical care.