Editors' ChoiceCancer

A 3D view of tumor heterogeneity

See allHide authors and affiliations

Science Translational Medicine  15 Nov 2017:
Vol. 9, Issue 416, eaaq1234
DOI: 10.1126/scitranslmed.aaq1234

Abstract

Solid tumor clearing and light-sheet microscopy improves interrogation of intratumoral heterogeneity.

Individual tumors can be staggeringly complex. Genome sequencing studies have identified extensive genetic heterogeneity between tumor cells. Even genetically identical cells can represent different phenotypes depending on the tumor microenvironment. Understanding intratumor heterogeneity could potentially inform how cancers are diagnosed and treated. However, the conventional needle biopsy method examines only a tiny fraction of a tumor in two dimensions, potentially missing important disease features. Genome sequencing methods require dissociating the tumors to form cell suspensions, causing a loss of vital environmental and spatial features.

In a recent study, Tanaka et al. developed a pipeline for three-dimensional (3D) imaging and phenotyping of whole human solid tumors. They combined an organic solvent-based tissue clearing method with light sheet microscopy to optically clear intact tumor tissues and image the entire volume at single-cell resolution. Using a pipeline they called DIPCO (diagnosing immunolabeled paraffin-embedded cleared organs), the authors were able to label intact formalin-fixed paraffin-embedded tumor tissues with diagnostic antibodies and visualize their 3D distributions. The obtained volumetric data allowed identification of hyper- and hypovascular niches and quantification of the heterogeneous distribution of various immunosignals.

Characterization of the spatial and heterogeneous features of CD34 signal (a marker of blood vessels) using the DIPCO pipeline provided more accurate information than did standard two-dimensional (2D) histological methods for various human solid tumors. In particular, when the DIPCO pipeline was applied to 3D analysis of CD34 expression in advanced ovarian cancer, it revealed an association between drug responsiveness and vessel radius, as well as CD34 density variance, which cannot be accurately measured by 2D histological methods.

This study demonstrates that 3D pathology approaches can have great potential to improve diagnostic staging of tumor samples and optimization of patient treatments. Unfortunately, the method presented here requires a long processing time (over 10 days), owing to the slow nature of passive tissue clearing and labeling. The development of more time-effective techniques will accelerate the adoption of 3D pathology approaches in the clinical setting.

Highlighted Article

View Abstract

Navigate This Article