Developing Novel Technology for Data Driven Clinical Decision Making
Lead Research Organisation:
Imperial College London
Department Name: Metabolism, Digestion and Reproduction
Abstract
Glioblastoma (GBM) brain tumours are deadly and incurable. Effective treatment options for GBM are urgently needed. The extent of surgical GBM tumour resection substantially positively impacts patient survival, but damage to normal brain tissue can severely affect quality of life and hence must be minimised. Therefore, it is imperative that we develop tools for surgical removal which can:
A) precisely remove cancer cells, with reduced thermal impact on normal brain cells
B) give feedback at the time of resection to surgeons so that they know which areas of the brain tissue are malignant and need to be removed and
C) give doctors indications of the nature of the cancer so that the most effective therapy can be employed post-operatively for any cancer cells left behind.
Development of this technology would be instrumental to revolutionising the way surgeons treat cancer. To do this we will develop high-precision lasers and understand their effects on tissue using very high-definition cameras, we will then collect the 'smoke' produced by these lasers when cutting and feed it into a mass spectrometer which is capable of measuring all of the individual metabolite chemicals representative of the tissue it has come from. When cancer is present, cells metabolise differently and hence this can be measured using the technology described. Observing the changes in metabolism can help us to understand complex diseases such as cancer and the therapies which may be effective against them. Therefore, in this project we will take the data from our novel technology and use machine learning to understand changes in the chemical compositions of the tissues and how that correlates to the different subtypes of cancer and use this information to predict the nature of any recurrent disease. If we can predict the recurrent nature of disease from the primary tissue, we may be able to use therapies more effectively to stop recurrence altogether.
A) precisely remove cancer cells, with reduced thermal impact on normal brain cells
B) give feedback at the time of resection to surgeons so that they know which areas of the brain tissue are malignant and need to be removed and
C) give doctors indications of the nature of the cancer so that the most effective therapy can be employed post-operatively for any cancer cells left behind.
Development of this technology would be instrumental to revolutionising the way surgeons treat cancer. To do this we will develop high-precision lasers and understand their effects on tissue using very high-definition cameras, we will then collect the 'smoke' produced by these lasers when cutting and feed it into a mass spectrometer which is capable of measuring all of the individual metabolite chemicals representative of the tissue it has come from. When cancer is present, cells metabolise differently and hence this can be measured using the technology described. Observing the changes in metabolism can help us to understand complex diseases such as cancer and the therapies which may be effective against them. Therefore, in this project we will take the data from our novel technology and use machine learning to understand changes in the chemical compositions of the tissues and how that correlates to the different subtypes of cancer and use this information to predict the nature of any recurrent disease. If we can predict the recurrent nature of disease from the primary tissue, we may be able to use therapies more effectively to stop recurrence altogether.
Technical Summary
Glioblastoma (GBM) brain tumours are deadly and incurable. Effective treatment options for GBM are urgently needed. The extent of surgical GBM tumour resection substantially positively impacts patient survival, but damage to normal brain tissue can severely affect quality of life and hence must be minimised. Therefore, it is imperative that we develop tools for surgical removal which can:
A) precisely remove cancer cells, with reduced thermal impact on normal brain cells
B) give feedback at the time of resection to surgeons so that they know which areas of the brain tissue are malignant and need to be removed and
C) give doctors indications of the nature of the cancer so that the most effective therapy can be employed post-operatively for any cancer cells left behind.
To do this we must understand the underlying biology and the mechanism of recurrence, GBM treatment resistance is not conferred by Darwinian-like selection of somatic mutations. Instead, two 'responder' subtypes have been identified based on therapy-driven changes in gene expression: all GBM tumours dysregulate the expression of a specific set of genes in response to standard treatment but in a patient-specific direction. We have coined these patients 'Up' and 'Down' responders. The biology of the genes involved indicates subtype-specific adaptive treatment resistance mechanisms. We would like to develop and use metabolic profiling technology; Laser Desorption Rapid Evaporative Ionisation Mass Spectrometry (LD-REIMS) to identify spatially resolved metabolic signatures associated with the responder subtype and be able to predict response from the primary tissues. By understanding the laser-tissue interaction we believe we will be able to generate novel technological advancement to progress biological understanding of this disease and embark on the pathway to clinical translation into a revolutionary surgical tool.
A) precisely remove cancer cells, with reduced thermal impact on normal brain cells
B) give feedback at the time of resection to surgeons so that they know which areas of the brain tissue are malignant and need to be removed and
C) give doctors indications of the nature of the cancer so that the most effective therapy can be employed post-operatively for any cancer cells left behind.
To do this we must understand the underlying biology and the mechanism of recurrence, GBM treatment resistance is not conferred by Darwinian-like selection of somatic mutations. Instead, two 'responder' subtypes have been identified based on therapy-driven changes in gene expression: all GBM tumours dysregulate the expression of a specific set of genes in response to standard treatment but in a patient-specific direction. We have coined these patients 'Up' and 'Down' responders. The biology of the genes involved indicates subtype-specific adaptive treatment resistance mechanisms. We would like to develop and use metabolic profiling technology; Laser Desorption Rapid Evaporative Ionisation Mass Spectrometry (LD-REIMS) to identify spatially resolved metabolic signatures associated with the responder subtype and be able to predict response from the primary tissues. By understanding the laser-tissue interaction we believe we will be able to generate novel technological advancement to progress biological understanding of this disease and embark on the pathway to clinical translation into a revolutionary surgical tool.
