Toward a quantitative method for estimating tumour-stroma ratio in breast cancer using polarized light microscopy

Abstract

The tumour-stroma ratio (TSR) has been explored as a useful source of prognostic information in various cancers, including colorectal, breast, and gastric. Despite research showing potential prognostic utility, its uptake into the clinic has been limited, in part due to challenges associated with subjectivity, reproducibility, and quantification. We have recently proposed a simple, robust, and quantifiable high-contrast method of imaging intra- and peri-tumoural stroma based on polarized light microscopy. Here we report on its use to quantify TSR in human breast cancer using unstained slides from 40 patient samples of invasive ductal carcinoma (IDC). Polarimetric results based on a stromal abundance metric correlated well with pathology designations, showing a statistically significant difference between high- and low-stroma samples as scored by two clinical pathologists. The described polarized light imaging methodology shows promise for use as a quantitative, automatic, and standardizable tool for quantifying TSR, potentially addressing some of the challenges associated with its current estimation.

Polarimetric biomarkers of peri-tumoral stroma can correlate with 5-year survival in patients with left-sided colorectal cancer

“…We propose a polarimetry-based unsupervised clustering pipeline for categorizing the likelihood a patient is to survive after 5 years, using polarimetric imaging, image analysis and machine learning to achieve encouraging initial results. This expands our recently developed methodology 34 – 37 by introducing GLCM texture features, implementing unsupervised machine learning clustering algorithms, and comparing our resultant cluster assignments directly with clinical data. This last point is particularly significant, as previous studies were largely correlated to pathology scoring and other (approximate and somewhat subjective) diagnostic/prognostic measures 34 – 37 .…”

“…This expands our recently developed methodology 34 – 37 by introducing GLCM texture features, implementing unsupervised machine learning clustering algorithms, and comparing our resultant cluster assignments directly with clinical data. This last point is particularly significant, as previous studies were largely correlated to pathology scoring and other (approximate and somewhat subjective) diagnostic/prognostic measures 34 – 37 . Comparing our predictions to actual patient survival outcomes is thus much more meaningful in the clinical context.…”

“…For each case, a representative unstained slide with the deepest tumour invasion area was cut from the formalin-fixed and paraffin-embedded blocks to 4.5 μm thickness. Sample preparation involved chemical dewaxing to avoid possible polarization imaging artefacts 34 – 37 . No further processing was required for polarimetric imaging.…”

“…3 ). The choice of ROI size was guided by previous studies, compromising between the need for robust statistics and capturing stromal heterogeneity while also maintaining reasonable spatial resolution/information 34 , 35 , 37 . The pathologist was blinded to the polarimetry images and clinical outcome data (i.e., 5-year survival) when choosing ROIs.…”

“…No staining or contrast agent is required. Imaging for this study was done using our group’s novel polarimetric methodology previously described in detail 34 – 37 . Briefly, we used an AxioZoom V16 microscope (Zeiss, Germany) fitted with two linear polarizers (Thorlabs, LPVISE100-A) on computer controlled motorized rotation mounts (PRM1.MZ8, Thorlabs, USA).…”