A multivariate analysis for predicting cisplatin-induced delayed emesis.

The present study was designed to address the question of the early identification of patients at risk of developing cisplatin (CP)-related delayed emesis. This study included demographic, clinical, biological and pharmacological data and was conducted on 110 consecutive patients treated by CP-based chemotherapy. A previously validated single-point CP pharmacokinetic evaluation was performed in all patients. A total of 110 cycles was analyzed. Delayed vomiting (i.e., occurring between day 3 and day 7 following CP administration) was observed in 36.4% of cycles. Among the tested variables, the occurrence of delayed emesis was significantly related to elevated ultrafilterable (UF) platinum concentration (measured 16 h after the end of CP administration) and to low plasma magnesium concentration (measured 48 h before CP administration). Risk-thresholds for delayed emesis were established for UF platinum and magnesemia, at 60 ng/ml and 58 mmol/l, respectively. In the sub-group of patients with magnesemia determination, this later parameter was the only significant predictor of delayed emesis. Gender, cycle number, primary cancer location and age were not associated with the risk of developing delayed emesis. The ability to select patients at risk of delayed vomiting may offer a practical means of targeting administration of specific treatments.