Increased sampling will lead to an increase in detection, but is it clinically relevant?

However, we would like to expressa great concern regarding the pathology protocols,which were used. The authors claim to have used theEuropean Organisation for Research and Treatment ofCancer (EORTC) Melanoma Group (MG) protocol.However there is an internal contradiction between theabstract and methods in their description of this proto-col and furthermore neither of the methods describedadequately reflects the EORTC protocol. There appearsto be a misunderstanding between the meaning of stepsand levels. A step is the distance between levels, a level isa section or sections stained to represent that level. TheEORTC protocol as it has evolved consists of six pairsof sections, each stained with haematoxylin–eosin andS100. These are separated by steps at 50 lm thickness,so there are five steps and six levels whereas in theabstract there is a description of five sections 50 m apartand in the methods there is a description of six levels50 m apart. It was then stated that the first five levelswere treated differently. Furthermore in Fig. 1 there isagain confusion between the words steps and sectionsor levels.Despite this confusion we would like to acknowledgethat increased sampling will lead to increased detectionrates, as it was also demonstrated by the study by Cooket al. The extensive protocol identified 33.8% versus 25.2for the EORTC MG protocol, this is close to the detec-tion rate achieved by RT-PCR.