The KRAS variant as a predictive biomarker of cetuximab response in head and neck cancer
A partir de données portant sur 891 patients atteints d'un carcinome épidermoïde de la tête et du cou de stade avancé et inclus dans un essai de phase III évaluant l'efficacité de l'ajout du cétuximab à un traitement combinant cisplatine et radiothérapie, cette étude analyse l'association entre un variant du gène KRAS et la réponse au traitement
In their interesting report, Weidhaas et al1 posit that a germline mutation in the 3′ untranslated region of the KRAS variant is a predictive marker of improved response to cetuximab for patients with locally advanced head and neck squamous cell carcinoma (HNSCC). While the results presented therein are intriguing, we would advise caution regarding their interpretation. The study by Weidhaas et al has several strengths; first, being focused on a well-characterized germline polymorphism that has been previously described as a predictive biomarker in a similar context for colon cancer.2 The evaluation of a single marker increases statistical power by limiting the number of comparisons performed3; however, while only 1 variant was analyzed, multiple hypotheses were evaluated, thereby increasing the potential for inflated type I error.3
JAMA Oncology , commentaire, 2015