232 Chapter 8 the assessment whether an END is necessary during the initial surgery. For OSCC with DOI ≤ 4 mm, D1C2 uniform positivity could aid in the clinical decision whether regular follow-up, watchful waiting, or END is more appropriate (chapter 6). Finally, it is hypothesized that MET-EC- contributes to the invasive nature of OSCC and that the low success rate of MET targeted therapies might be explained by a lack of CDx that account for ectodomain shedding (chapter 7). Below, the results obtained in this thesis will be discussed in the context of questions that arose while compiling this manuscript, ultimately leading to the proposed theorems regarding this work. Was D1C2 the most reliable choice? The reliability of five C-terminal MET antibodies was investigated using the Rimm Lab Algorithm for antibody validation (23) as guidance (chapter 3), before examining the prognostic and potential predictive value of membranous MET immunoreactivity (chapters 3 -5, and 7). The results illustrate that only two out of five investigated antibodies, D1C2 and CVD13, reliably detect complete MET and/or MET C-terminal fragments under all examined conditions (i.e. reducing, native, and FFPE) confirming the need for antibody validation. Since D1C2 was most sensitive in the detection of membranous MET under FFPE conditions, this clone was used for all C-terminal MET immunohistochemical analyses performed in this thesis. In retrospect, some nuances are in place. The original antibody validation Rimm Lab Algorithm for immunohistochemistry and quantitative immunofluorescence (23) used ‘western blotting a panel of cell lines selected for the target of interest’ as a starting point. However, in a more recent publication by the hand of the same group (24), specifically addressing immunohistochemistry, used ‘examination of the expected cellular localization in a tissue of interest using immunohistochemistry’ as the starting point of the algorithm. It is only in step 3, after titration (step 2), of the updated algorithm that they advise to perform an orthogonal method, e.g. western blot, or mass spectroscopy. Furthermore, it should be mentioned that western blotting has its limitations as an orthogonal validation method. Since western blotting is often performed under denatured conditions, it does not imply corresponding results with immunohistochemistry in which the antigen/protein is in its native form. Supposedly, having followed the more recent flow chart, never incorporating western blotting as an orthogonal method for validation, the reported unspecific 90 kDa band detected with SP44 in SK-BR-3, HT-29 and LNCaP could have gone undetected, possibly leading to the choice of SP44 as the most reliable (sensitive and specific) antibody in the immunohistochemical detection of membranous C-terminal MET. As we have never performed western blotting of
RkJQdWJsaXNoZXIy MTk4NDMw