The rate of RET rearrangement in NSCLC was as low as 0.7%, while previous studies have reported frequencies up to 6%. RET rearrangements have been suggested to be more frequent in Asian patients, in non-smoker and in the adenocarcinoma subgroup. Therefore, the selection of the study population might significantly affect the frequency at which RET rearrangements are detected. In addition, RET rearrangement is usually mutually exclusive to other driver mutations, and its frequency results higher in patients that do not harbor the more frequent KRAS and EGFR mutations [
3]. Nevertheless, this observation poses a major problem for molecular diagnostics that is common to many cancer types. In fact, the number of potential predictive biomarkers that might offer possibility of therapeutic intervention in lung cancer as well as in other tumor types is increasing exponentially (Table
1). Identification of driver mutations might result in a survival advantage for cancer patients that have access to novel drugs through clinical trials or, in selected cases, to receive an off-label treatment with agents approved for other indications [
9]. However, the time, the cost, and the amount of tissue needed for a wide molecular profiling using routine diagnostic methods are not compatible with the standard clinical workup, in particular in lung cancer. In many European countries diagnosis of lung cancer is based in over 50% of the cases on cytology samples or small biopsies that might not be sufficient for analysis of somatic mutations and gene rearrangements in several different genes using sequencing, Real Time PCR and/or FISH. Indeed, a 26.9% failure rate in FISH analysis due to an inadequate number of tumor cells or sample quality was reported by Platt and colleagues [
8]. This observation underlines the need for novel methods in molecular diagnostics that allow a comprehensive molecular characterization of lung tumors in the routine clinical workout [
10]. In this respect, different methods to detect mutations and fusions using genotyping or targeted next generation sequencing are being explored, and might be ready in a short timeframe for clinical implementation [
11-
14].