We read with interest the study by Potrony et al. reporting no influence of the age of formalin-fixed and paraffinembedded (FFPE) melanoma samples on BRAF and NRAS sequencing. In their study, Potrony et al. encountered a DNA extraction failure in 10% of melanoma samples [1]. This rate of inconclusive samples is also encountered in other studies dedicated to mutational screening in melanoma samples [2,3].In our experience, we recently lead two successive studies to evaluate BRAF and NRAS mutational status using pyrosequencing and BRAFV600E and NRASQ61R mutation specific immunohistochemistry in melanoma samples. We analyzed, on the one hand, retrospectively 142 FFPE samples from 2003 to 2013 [4]. On the other hand, in 2014, 111 samples were analyzed in a prospective study [5]. In the retrospective study, the rate of inconclusive molecular analyses was 12.3% whereas in the prospective one only 5.9% of the molecular analyses remained inconclusive (p = 0.02, Student's t-test) [4,5]. There was a significant correlation between the age of the FFPE samples and the success of molecular analyses, the older samples being often inconclusive about BRAF and/or NRAS mutational status (correlation coefficient r = 0.4913, p < 0.0001) [4,5]. Rather than the time of fixation and the storage conditions of the FFPE samples which were almost homogenous, we hypothesized that the main factor causing the failure of DNA analyses was an inadequate fixation process using ancient non standardized and various fixative process as, for example, sodium acetate acetic acid formalin or Bouin's fixative in some of the oldest samples. With the generalization of the 4% buffered neutral formalin fixation, the rate of inconclusive DNA analyses actually remains stable in our daily practice at about 6%-7% of the analyses performed in melanoma samples. Additional criteria as low percentages of tumor cells in some samples are also partially responsible of these inconclusive analyses but a few samples still do not allow correct DNA extraction. Mutation specific immunohistochemistry could help to interpret these challenging and molecularly inconclusive samples searching for the mutant proteins instead of the mutations at the DNA level [2][3][4][5][6].In this way, we agree with Potrony et al. that the time of conservation of the FFPE samples does not greatly influence the success of DNA analyses but we also want to remind the reader that the fixation process is a crucial step in the management of melanoma and, in a larger way, of every solid tumor sample to allow further molecular analyses. Four percent buffered neutral formalin fixation during about 24 h seems to be an adequate fixation process to allow histopathological, immunohistochemical and also additional DNA-sequencing analyses. Standardized fixation and conservation process are crucial to permit contributive new analyses on FFPE tumor samples. This will permit updating of the diagnoses of tumors and to improve the therapeutic management of the patients taking into account the growing and evolving fie...