Electrochemotherapy as palliative treatment in patients with recurrent and/or metastatic head and neck tumours: features analysis for an early determination of the partial responsive patients.

OBJECTIVE: The aim of this study was to identify features mainly involved in determining the partial response (PR) to the Electrochemotherapy (ECT) in patients with recurrent and/or metastatic head and neck (H&N) tumor; the identified features were also used in a decision chart in order to provide the clinician with a support tool in deciding further therapies. PATIENTS AND METHODS: 131 patients (186 treatment sessions) with recurrent and/or metastatic H&N neoplasm were subjected to ECT. Treatment response was evaluated based on Response Evaluation Criteria in Solid Tumors (RECIST) v. 1.1 two months after the ECT. The grade of bleeding and pain before, at the end and one week after ECT treatment were evaluated. Univariate and multivariate analysis were performed to identify features involved in determining the patient PR. RESULTS: In the context of the univariate analysis, tumor size significantly influenced the response to ECT, with higher PR rate of 58.3%: 28 among 48 patients with lesion size ≤ 3 centimeters (p-value < 0.001 at Chi-square test). Pain and bleeding pre-treatment were positively correlated to PR (p-value < 0.001 at Chi-square test). A difference in the current flowing in the tissue during treatment was also observed in partially responsive patients, where the median current value (6.6 A) was higher than that achieved in patients that did not show PR (3.3 A). In the context of the multivariate analysis, the best performances are achieved with the BART method (accuracy of 84%). The main clinical factors to predict the partial response, among investigated features, that have shown to be considered were the pain value felt before performing the treatment and the median current delivered during the ECT treatment. A decision-making support tool to predict the patient prognosis in terms of response rate could be represented by the decision tree obtained with CART algorithm, where a pain pre-treatment more than 5 and a median delivered current not less than 2.8 A led to the prediction a partial responsive patient with an accuracy of 75%. CONCLUSIONS: The study confirmed that ECT is an interesting antitumoral therapy in advanced chemo- and radio-refractory H&N neoplasms, able to reduce frequent symptoms and to improve the quality of life. Pain pre-treatment and delivered current are the most important variables when predicting the partial response of patients.