A New Prognostic Instrument for Predicting the Probability of Completion of Cisplatin during Chemoradiation for Head and Neck Cancer.

Abstract

Many head and neck cancer patients assigned to definitive or adjuvant chemoradiation treatment do not complete the concurrent cisplatin dose. We determined corresponding risk factors and developed a prognostic instrument to help identify these patients. Ten pre-treatment characteristics were retrospectively analyzed in 154 patients with head and neck cancer who were treated via chemoradiation with cisplatin. These pre-treatment characteristics included age, sex, Karnofsky performance score, tumor site, primary tumor stage, nodal stage, histologic grade, upfront surgery, human papilloma virus status, and history of smoking. The characteristics significantly associated with the completion of cisplatin-based treatment, the receipt of ≥80% cisplatin, or showing a strong trend of association after multivariate analyses were used for the prognostic instrument. For each characteristic, 0 points were assigned for worse outcomes, and 1 point was assigned for better outcomes. Patients' scores were calculated by adding these points. Age ≤ 60 years and a Karnofsky performance score of 90-100 were significantly associated with both endpoints after multivariate analysis, and male gender showed a trend for association with the receipt of ≥80% cisplatin. Patient scores were 0, 1, 2, and 3 points. The corresponding rates of completion of cisplatin-based treatment were 14%, 41%, 62%, and 72%, respectively (p = 0.004). The rates of receipt of ≥80% cisplatin were 29%, 54%, 72%, and 94%, respectively (p < 0.001). This new prognostic instrument helps to predict whether head and neck cancer patients scheduled for chemoradiation will receive cisplatin as planned.