: Molecular Classification of Tumours by Spectroscopic Analysis of Tissue Replicas on Nanoneedles

Student thesis: Doctoral ThesisDoctor of Philosophy


Precision oncology aims to predict the optimal therapy for each cancer patient, and one of the main resources to tackle this challenge is molecular diagnostics. Molecular diagnostics promises to tailor treatments to the individual characteristics of each malignancy by providing molecular-level and mechanistic understanding of the nature of individual pathologies and response to treatment. Yet, it largely relies on biomarker detection, which requires labour intensive protocols and provides limited information on the overall molecular signature of tumours and their heterogeneity. Currently, molecular diagnostics is going through a shift beyond the use of few panels of biomarkers, to better capture the whole molecular signature of the tumours and improve prognosis and treatment selection. For tumours that are difficult to reach such as glioma, another limitation resides in the invasiveness of the biopsy collection process. This limits the number of tissue biopsies which can be obtained without endangering health of the patient. Overcoming these limitations can improve the therapeutic efficacy of molecular diagnostics for the treatment of gliomas.

Nanobiopsy develops a platform for accurate, minimally invasive, label-free mapping of the molecular profile of a tumour. This map allows stratifying tumours based on both molecular composition and spatial distribution (heterogeneity). Ordered arrays of conical-shaped porous silicon nanoneedles (nNs) are fabricated and pressed onto the tissue, to harvest molecules with little perturbance of cell activity. The in vivo safety of nNs has been previously assessed, along with their ability to probe the intracellular environment. Using nNs to generate a molecular replica of brain tissue, Nanobiopsy can reduce the need for biopsies, significantly reducing invasiveness, and thus increasing sampling frequency and extent. NANOBIOPSY analyses replicas by Raman spectroscopy imaging and desorption electrospray ionization (DESI) mass spectrometry imaging, to collect maps of the tissue molecular fingerprints. Multivariate analysis can stratify clinical samples according to the shared molecular profile.

I optimised the parameters for the fabrication of nNs, that are suitable for the harvesting of biomolecules in Nanobiopsy. Adapting standard extraction, purification and total quantification methods, I observed that nNs can harvest RNA, proteins and lipids in comparable amount to that available within the tissue. Raman and DESI-MS images of replicas and proximal sections (as a control) from tumour-bearing murine brains were obtained and analysed with unsupervised multivariate analysis algorithms. The analysis showed the capability of Nanobiopsy to discriminate healthy tissues and glioma both on sections and nNs. Using the same approach and DESI-MS as imaging technique, graded human biopsies from glioma patients were analysed to find relevant spectral features for the stratification of the malignancy.

In conclusion, I showed that nNs can be tailored to harvest molecules that replicate tissue-specific compositional and morphological information. Label-free spectral techniques can be used to image the replica, enabling NANOBIOPSY to detect gliomas based on a whole molecular fingerprint. This can help addressing the precision oncology challenge of developing effective strategies for molecular diagnostics that can contribute to a personalized patient treatment.
Date of Award1 Aug 2022
Original languageEnglish
Awarding Institution
  • King's College London
SupervisorCiro Chiappini (Supervisor), David Richards (Supervisor) & Mads Bergholt (Supervisor)

Cite this