Optimal cumulative cisplatin dose in nasopharyngeal carcinoma patients based on induction chemotherapy response

Nasopharyngeal carcinoma (NPC) patients can be separated into two risk subgroups according to tumor responses to induction chemotherapy (IC). We aimed to elucidate the optimal cumulative cisplatin dose (CCD) of concurrent chemoradiotherapy (CCRT) for different NPC patient subgroups.A total of 990 patients with incident NPC diagnosed between 2008 and 2017 treated with IC plus CCRT were included in our observational study. The clinicopathological features of patients with different tumor responses were compared using the Chi-square test or Fisher's exact test. Prognosis was assessed using a multivariate Cox proportional hazards model. In addition, acute and late toxicities were compared between different CCD groups.After IC, 761/990 (76.9%) patients had a complete tumor response (CR)/partial response (PR) and 229 (23.1%) had stable disease (SD)/disease progression (PD). An unsatisfactory tumor response (SD/PD) after IC correlated with poor clinical outcome (3-year PFS 61.4% vs. 83.2%, P < 0.001 and 3-year LRFS 80.9% vs. 94.5%, P < 0.001). Patients who achieved CR/PR after IC received a CCD >200 mg/m2 and showed higher 3-year PFS and DMFS rates than those receiving a CCD <100 mg/m2 (PFS: 85.4% vs. 77.9%, P = 0.045; DMFS: 89.4% vs. 77.9%, P = 0.015). Multivariate analysis also showed that CCD was an independent prognostic factor for PFS and DMFS in CR/PR subgroup. Moreover, the medium dose group showed similar efficacy as high dose group but was associated with fewer grade 1-4 acute toxicities. However, application of different CCD didn't result in significantly different survival outcomes in SD/PD subgroup.Tumor response to IC was an independent prognostic factor for patients with NPC. For the patients who achieved CR/PR after IC, patients receiving high CCD showed significantly improved 3-year PFS and DMFS compared with patients receiving low CCD. Balancing toxicity and efficacy, 200 mg/m2 seemed to be the optimal dose in the CR/PR groups. However, enhancement of CCD did not provide survival benefit for patients who achieved SD/PD after IC, and treatment options for these patients require further consideration.