The CONDOR study, a randomized phase II study, investigated feasibility of docetaxel/cisplatin/5-fluorouracil (TPF) followed by conventional radiotherapy with cisplatin 100 mg/m2 on days 1, 22, 43 (cis100 + RT) versus accelerated radiotherapy with cisplatin weekly 40 mg/m2(cis40 + ART) in LAHNC pts. The effect of the two regimens on ototoxicity was investigated.
Audiometry was carried out at baseline, during and after TPF before start of CRT, and 1, 4, 8 and 12 months after end of treatment. Air-conduction thresholds were determined in octave steps from 1 kHz until 8 kHz. Where 1 to 4 kHz is relevant for speech. Based on baseline thresholds we divided the pts in two groups; pts with baseline thresholds 50 dB.
62 pts started with TPF; 56 pts were randomized to cis100 + RT (n = 27) or cis40 + ART (n = 29). Compliance to audiometry was low, mostly due to poor physical condition. Pts included in this analysis were 12 in cis100 + RT and 11 in cis40 + ART. Mean cumulative cisplatin dose was 498 mg/m2 (SD 66.1) for cis100 + RT and 490 mg/m2 (SD 55.8) for cis40 + ART (p = 0.75). Hearing deterioration over time was gradually for cis40 + ART and abrupt for cis100 + RT; with a wide spread in both groups. Mostly, the abrupt hearing deterioration occurred during CRT. Hearing loss was most prominent at 8 kHz and almost absent at 2 kHz. We used the Wilcoxon test for our hypothesis that pts treated with cis100 + RT suffer more hearing loss than pts treated with cis40 + ART; showing a difference for 8 kHz (z = 2.07; p = 0.03) and 4 kHz (z = 1.98, p = 0.04). Analysis for hearing deterioration in the subgroup of baseline hearing 50 dB mean deterioration was less, approximately 10 dB. In 15 pts we could compare data from 12 versus 4 months after end of treatment. These showed no clinically relevant at 8 kHz.
|Frequency||Cis100 + RT||Cis40 + ART|
|8 kHz||41 dB||20 dB|
|4 kHz||30 dB||12 dB|
|2 kHz||7 dB||3 dB|
After TPF CRT with cis40 + ART is less ototoxic than CRT with cis100 + RT.
Clinical trial identification
Legal entity responsible for the study
Carla van Herpen
Sanovi Aventis Netherlands Dutch Cancer Society
All authors have declared no conflicts of interest.