Individualised cumulative cisplatin dose for locoregionally advanced nasopharyngeal carcinoma patients based on induction chemotherapy response and tumour volume.

Background and objectives: To evaluate the prognostic value of an integrated model consisting of tumour response to induction chemotherapy (IC) and gross tumour volume (GTV) after IC in nasopharyngeal carcinoma (NPC) and elucidate optimal cumulative cisplatin dose (CCD) in concurrent chemoradiotherapy (CCRT) for different subgroups.

Design and methods: This retrospective study enrolled 896 patients with NPC diagnosed from 2010 to 2017 receiving IC plus radiotherapy. The primary endpoint was disease-free survival (DFS). Cut-off points for GTV were combined with IC response to develop an integrated model. Propensity score matching (PSM) was used to adjust for potential confounders. Survival outcomes and acute toxicity were compared between the different CCD groups.

Results: Unsatisfactory IC response and large GTV after IC were correlated with poor survival outcomes; the AUC increased to 0.668 when these factors were incorporated. The integrated model classified patients into three groups. After PSM, radiotherapy alone and CCRT demonstrated similar efficacy in the low-risk group (complete response (CR)/partial response (PR) and GTV <68 cm³ after IC). In the intermediate-risk group (CR/PR but GTV ⩾68 cm³), CCD of >200 mg/m² and 101-200 mg/m² increased the 5-year DFS rates (83.7% vs 81.1% vs 65.3%, p = 0.042). In the high-risk group (stable disease/progressive disease and any GTV), the use of different CCDs did not result in significantly different survival outcomes (p = 0.793). Additionally, high CCD was significantly associated with increased incidence of grade 1-4 acute toxicity.

Conclusion: The integrated model incorporating IC response and GTV after IC demonstrates satisfactory value in risk stratification and the potential to guide individualised decision-making in CCD selection. Balancing toxicity and efficacy, RT alone seems to be the optimal treatment for patients in low-risk groups and 200 mg/m² might be the optimal dose for intermediate-risk groups. Moreover, increasing CCD does not benefit patients in high-risk groups, and treatment options for these patients require further consideration.