Risk prediction model for cisplatin-induced acute kidney injury in patients with head and neck cancer receiving chemoradiotherapy: A re-analysis of a phase II/III JCOG1008 trial.

OBJECTIVES:Acute kidney injury (AKI) represents a major toxicity associated with cisplatin. We developed a risk prediction model for cisplatin-induced AKI in patients with postoperative high-risk head and neck cancer who received chemoradiotherapy during a randomized phase II/III trial, JCOG1008. MATERIALS AND METHODS:Two hundred and fifty-one patients received radiotherapy with weekly cisplatin at 40 mg/m2 (weekly arm) or 3-weekly cisplatin at 100 mg/m2 (3-weekly arm). AKI was defined using the AKI Network classification/staging system as increased serum creatinine of ≥0.3 mg/dL or a ≥1.5-fold increase from baseline 30 days after completing chemoradiotherapy. The Akaike information criterion was used to explore the optimal model by combining explanatory variables at registration. RESULTS:Among the 251 patients (210 men and 41 women (median age; 62 years)), 94 (37.5 %) developed cisplatin-induced AKI. The optimal cisplatin-induced AKI risk prediction model comprised four factors, including a primary site of hypopharynx/larynx (vs. oral cavity/oropharynx), 3-weekly arm (vs. weekly arm), serum albumin of ≤3.5 g/dL (vs. >3.5 g/dL) and creatinine clearance (CCr) of <90 mL/min (vs. ≥90 mL/min). The incidence of cisplatin-induced AKI rose with cumulative count of the four factors. When the cumulative count was ≥2, the positive predictive value for cisplatin-induced AKI was 50.3 %. CONCLUSIONS:We developed a risk prediction model for cisplatin-induced AKI in patients with head and neck cancer who received postoperative chemoradiotherapy using primary site, cisplatin administration method, serum albumin, and CCr. Patients with risk factors unrelated to the cisplatin administration method should adopt a weekly cisplatin regimen.