Next-generation Pathology

Pathology faces increasing competition from the modern methods of molecular biology. Genomics and proteomics promise to provide unbiased, quantitative data revealing insights into origin and progression of disease. Will pathology need to change in order to defend its territory? Quite certainly. Should pathologists forget what they have learned over the centuries? I don’t think so.

It could be claimed that the basic pathological methods have changed too little since the invention of the light microscope. However, the basic reason to look at histological samples didn’t change either. The morphological details of heterogeneous tissue provided invaluable information well before genes and proteins were discovered. And they continue doing so ever since.

Genes and proteins are essential components of the molecular magic that brings about tissue morphology (among other things). No biological science can afford ignoring all we have learned about them in the recent past. At the same time, genetic fingerprints and protein profiles are mostly obtained from blood or homogenized tissue samples that lack the spatial dimension histology is all about.

Tissue slides reveal the complex interplay between different cellular populations, their gene expression levels and metabolic status. They thus are a more comprehensive readout of molecular dynamics than any molecular technique alone could ever provide. Pathological scores, though, are necessarily less comprehensive. They aim at reducing complexity in order to increase consistency. They are so successful at achieving the former, that the quantitative requirements of modern science are hardly met. Yet they often fail to establish the latter as inter-observer variation remains high.

Next generation pathology will need to integrate molecular techniques into the traditional framework of histology: the morphological complexity of heterogeneous tissue. The required methods are well established and they are now transforming pathology. Immunohistochemistry combines molecular information on protein levels with spatial context. Novel methods based on in situ hybridization are reaching the maturity required to complement the histological toolbox. One by one, IHC stains put proteins into context. SISH, CISH, FISH et al. now do the same with genes. These advances open another dimension to the information that can be derived from histological samples. The spatial dimensions can be complemented with molecular ones. And the molecular information can be added to the information residing in tissue instead of competing with it.

As usual, the next generation comes along with novel toys. Automated image analysis is required to capture the quantitative readouts promised by the molecular advances. In order to go beyond genomics and proteomics, pathological knowledge will have to be accounted for and it is thus essential to make the novel approaches available to pathologists. To get from toys to tools, automated image analysis solutions need to be straightforward enough to be used routinely by pathologists. At the same time, they need to be flexible enough to robustly cope with the heterogeneity of histological samples.

Tissue proteomics and tissue genomics should be the pathologist’s reply to the molecular challenge they face.

For more information on this, check out Definiens Webinar series, including one this Thursday entitled "Quantitative Digital Pathology Out-of-the-Box: Definiens Tissue Studio Combines Flexibility and Ease-of-Use with High Productivity". Click here to register.

Thursday March 29th, 2012