Towards High-Throughput Detection of Ovarian Cancer Cells from Liquid Biopsy using Stimulated Raman Scattering

While the previous two decades have seen a dramatic improvement in general cancer survival rates, for some pervasive cancer types such as Ovarian Cancer (OC), survival rates have only marginally increased. This is mostly due to the late stage at which OC’s is typically diagnosed.

Detection of circulating tumour cells (CTCs) would present a promising opportunity for advancing OC diagnosis; however, blood CTCs are very scarce (20 per 10^9 cells). For all cancer types lacking a good biomarker (e.g. OC), a high-throughput screening technique for whole blood would revolutionize early diagnosis.

This proposal will form the foundation of a project to develop a label-free microfluidic screening platform that can search millilitres of whole blood for CTCs, using Stimulated Raman Scattering as the detection modality.

Because the resources required for this ambitious goal are expensive, this proposal will focus on the most critical development: creation of a scalable method for detecting and sorting of tumour cells.

A two-orders-of-magnitude increase in throughput will be achieved by performing the detection and sorting of CTCs in multiple microfluidic channels simultaneously. Once isolated, CTCs can be further analysed by sequencing, culturing or any other diagnostic technique.

To demonstrate proof-of-concept while simultaneously achieving research impact, initial studies will focus on the detection of tumour cells in ascites, rather than blood, using spontaneous Raman scattering. Tumour cells are much more prevalent in ascites but their isolation is still challenging.

Spontaneous Raman has already been used to identify tumour cells by using machine learning, and while it cannot scale to the throughputs required for blood screening, it should be sufficient for ascites.

Multiple microfluidic channels arranged side-by-side will be simultaneously excited by a single laser to identify tumour cells; Raman-scattered light will be captured by an imaging spectrometer, whereupon machine learning algorithms running on a Field Programmable Gate Array will identify the CTCs and sort them using laser cavitation.

The results of this research will support development of a 10,000x faster Stimulated Raman Scattering system able to perform label-free isolation of CTCs from whole blood.

This ability to quantify the number and type of tumour cells will find use throughout CRUK’s screening and early diagnosis remit, but also in staging, monitoring and research.

This study will kick-start Dr Sesen’s independent cancer research career, using his skills and knowledge of microfluidics and optics to achieve early diagnosis of multiple cancer types, as well as providing prognostic insight for physicians.