Tissue microarrays for immunohistochemical determination of oncological biomarkers

Abstract

Dear Editor, In a recent issue of Virchows Archiv, Marx et al. [1] investigated the potential prognostic impact of the expression of the biomarker alpha-methylacyl-CoA racemase (AMACR) detected in a large series of colorectal cancers (CRC) by immunohistochemistry (IHC) with tissue microarray (TMA) technology. Authors concluded that “AMACR cannot serve as a prognostic biomarker in CRC.” This conclusion was in contrast with the results obtained by Lin et al. [2] in a previous study where it was found that there was a striking prognostic role of AMACR detected by IHC using TMA in CRC. In [1], authors stated that “...it appears unlikely that the use of TMA containing only limited amounts of tumor tissue per patient (one spot of 0.6 mm diameter per patient) has led to a significant number of false negative cases.” and that “...focal AMACR expression cannot serve as explanation for the discrepancies” between the two studies as both were based on IHC determination of AMACR by TMA. Conversely, we think that the above discrepancies may be better explained by focusing on the potential technical pitfalls involved in TMA construction in the two studies rather than focusing on the data obtained. In fact, although IHC using TMA is substantially faster and at markedly lower costs compared with the conventional approach, its reliability and validity are influenced by many factors, the impact of which have to be put under control. For example, one of the major criticism of TMA technology is that it uses only a small fraction of a tissue specimen, which may not be representative of the whole tissue section. Although various studies [3, 4] have been conducted to optimize the sampling strategy for various solid cancers, actually there are no standardized operative procedures (SOP) for many biomarkers expressed in CRC, including AMACR. For example Marx et al. [1] used TMAs containing one core of 0.6 mm per specimen, whereas Lin et al. [2] used TMAs including three cores of 0.6 mm sampled from each specimen. Which is the minimum core numbers for specimen representativity need for this biomarker? Studies are needed to clarify this as well as many other aspects related to the TMA-based detection by IHC of AMACR in CRC. Among these, an important key issue concerns the antigen survival. It is important to clarify whether archival tissues retain their antigenicity despite long-time storage as paraffin blocks. TMA slides must be cut once and may not be used for staining until months or ever years later. On the other hand, although paraffin should protect the tissue from oxidation or other damage, some authors [5] have reported that, once tissue blocks are sectioned, the antigenicity of proteins on the slides would quickly degraded or even become lost with storage time, resulting in false-negative results. How many false-negative cases were expected for AMACR in [1] and [2] by considering the storage time of the specimens? Again, to answer this question, ad hoc studies are needed. In general, before clinically validating AMACR as prognostic biomarker in CRC, it should be necessary to validate the methodology (i.e., TMA technology) used for its detection. The workflow involved in the TMA validation should ideally include, as first step, studies aimed to define SOPs for each of the involved phases. In the second step, ad Virchows Arch (2009) 454:353–354 DOI 10.1007/s00428-009-0737-7