Statistical Model Predicts Cisplatin Chemoradiotherapy-Induced Hearing Loss

A prediction model that incorporates patient and treatment characteristics could identify head and neck cancer patients at risk of chemoradiotherapy-related hearing loss

medwireNews: A model based on patient and treatment features can be used to predict post-treatment hearing capacity in head and neck cancer patients given cisplatin -based concomitant chemoradiotherapy, say Dutch researchers.

Although the model needs to be validated externally, they write in JAMA Otolaryngology – Head & Neck Surgery, it nonetheless is “a step toward improving individual counseling of patients with head and neck cancer who are at risk for [concomitant chemoradiotherapy]-related hearing loss”.

The prediction model was based on chart data from 81 patients with advanced head and neck cancer treated with high-dose cisplatin chemoradiotherapy who received a total cumulative cisplatin dose ranging from 315 mg to 600 mg and a cochlear radiation dose between 1.1 Gy and 70.9 Gy.

The model incorporated the cisplatin and radiation dose together with pretreatment hearing levels, as assessed by pure-tone audiometry, to predict a post-treatment hearing level of at least 35 dB, which the study authors explain is the Dutch threshold for hearing aid qualification.

The model had an area under the receiver operating characteristic curve of 0.68, and a predicted post-treatment hearing threshold of at least 40 dB had 29% sensitivity and 97% for an observed hearing threshold of at least 35 dB, while the positive- and negative-predictive values were 78% and 76%, respectively.

When alternative cutoff values below 40 dB for predicted hearing levels were used, the sensitivity increased but the specificity decreased, with sensitivities ranging from 33% to 54% and specificities from 95% to 86% for cutoff values between 39 db and 35 dB, respectively.

However, the researchers say that they considered high specificity, and the corresponding low false-positive rate, as “most relevant” because a false-positive prediction could lead to an unnecessary switch to a less ototoxic treatment without “exact knowledge of its effect on tumor control”.

Lead author Eleonoor Theunissen, from The Netherlands Cancer Institute in Amsterdam, and co-investigators speculate that the low sensitivity indicates that “unknown variables are currently missing in the model, such as individual sensitivity to ototoxic effects”, adding that “future research concerning more variables as risk factors for hearing loss is needed”.

Furthermore, given that their model is based on head and neck cancer patients, the authors highlight the necessity of external validation other patient cohorts, such as cisplatin-treated patients with lung, bladder or gynaecological malignancies, as well as those given cisplatin in a postoperative setting.


Theunissen EAR, Zuur CL, Jóźwiak K, et al. Prediction of Hearing Loss Due to Cisplatin Chemoradiotherapy. JAMA Otolaryngol Head Neck Surg; Advance online publication 20 August 2015. doi:10.1001/jamaoto.2015.1515

medwireNews ( ) is an independent medical news service provided by Springer Healthcare Limited. © Springer Healthcare Ltd; 2015