Brillouin microscopy in cancer research: a review

Abstract

Significance: Cancer is one of the leading diseases worldwide, continuing to pose a significant financial burden to national health systems and taking lives. These drive the development of early-stage cancer diagnostics, which is believed to be a crucial step in improving patients' life expectancy and long-term outcomes of cancer treatments.

Aim: In this review, we explore the potential of a label-free technique known as Brillouin microscopy, a type of optical elastography, emerging as a promising candidate for early-stage cancer screening.

Approach: We discuss the main principles of this advanced imaging technology and provide a thorough analysis of all known Brillouin microscopy reports in application to cancer research and diagnostics. In our analysis, we focus on the mechanobiological aspect of the disease and draw conclusions based on four main sample types: cell cultures, cells in microfluidic environments, organoids, and excised tissues.

Results: We review recent advancements in cancer detection, finding that the technique can consistently biomechanically delineate between healthy and unhealthy cells, and organoids and tissues across multiple cancer types. We also present strides made in imaging mechanical changes in cancer during varying stages of progression, treatment, and regression.

Conclusions: We conclude this review with our perspective on the key developments required for technology's translation into the clinical realm, including measurement standardization, inclusion of statistical and artificial intelligence methods into data analysis and automated diagnosis, and further hardware developments needed for in situ and in vivo micromechanical measurements.

Keywords: Brillouin imaging; Brillouin microscopy; cancer detection; optical elastography.

Fig. 1

(a) Schematic of Brillouin light scattering within a biological sample made in Biorender. (b) Multiple spectra of varying common materials measured by a Brillouin Microscope, showing Brillouin Stokes and anti-Stokes peaks for several samples: water (light blue), stiff hydrogel (dark blue), and plexiglass (green).

Fig. 2

Timeline of Brillouin microscopy development constructed from a query of NIH’s PubMed database (gray) and a local collation of all known published Brillouin microscopy reports applied directly to cancer research (blue). The PubMed query searched for the terms “Brillouin imaging,” “Brillouin microscopy,” and “Brillouin microspectroscopy,” mentioning the technique anywhere within the paper.

Fig. 3

Types of apparatus for Brillouin microscopy: (a) a microscopy setup assisted with a tandem Fabry–Perot (TFP) spectrometer and (b) Brillouin microscopy based on virtually imaged phase array (VIPA) spectrometer. M, Mirror; L, Lens; BS, beam splitter; PBS, polarization beam splitter; FP, Fabry–Perot.

Fig. 4

Overview of various cancerous sample models analyzed using Brillouin imaging: (a) 2D cell cultures, (b) cell flow cytometry using microfluidics, (c) spheroids, and (d) animal or human tissues. All diagrams in Fig. 4 were made in Biorender.

Fig. 5

Biomechanical properties of cancer spheroids measured by Brillouin microscopy at 660 nm. (a) Brillouin and bright-field images of a normal spheroid (M1) and a cancerous spheroid (M3) acquired on days 2, 5, and 8. Scale bar: $10\mu \mathrm{m}$. Image reproduced from Hilai et al. (b) Average Brillouin frequency shifts of the core and periphery regions of M1 (top) and M3 (bottom) spheroids. Image reproduced from Hilai et al. (c) Brillouin frequency shift (GHz) at the core and periphery regions within spheroids samples (based on separate studies by Cheberkanov et al., Hilai et al. and Margueritat et al.⁵⁵).

Fig. 6

Graphics adapted from the study by Filipe et al. (a) Histological (H&E) staining of mouse murine mammary tumor brought under a Brillouin Microscope equipped with 660 nm laser source (scale bar = 1 mm). (b) Zoomed in 2D Brillouin frequency shift heatmap corresponding to the same surface region of the tumor shown in image (a). (c) Overlay of images (a) and (b) highlighting “soft” and “stiff” regions as identified by the 2D Brillouin frequency shift heatmap. (d) Brillouin frequency shift within a $2{\mathrm{mm}}^2$ section of the tumor, compared with the agar embedding material (**** = $p<0.0001$).