Development of multiphoton microscopes for real-world clinical applications

At present, the diagnosis of many types of disease must be confirmed by histopathology before treatment can begin. This entails the physical removal of a small tissue specimen from the patient that is then processed, sliced, stained and analysed by an expert using an optical microscope. The collection of tissue is subject to sampling error, i.e. the diseased tissue may be missed, and the time required for processing and analysis can delay treatment. It would desirable for physicians to be able to make an immediate diagnosis during examination of the patient.

When biological tissue is illuminated with light at appropriate wavelengths, certain naturally occurring biomolecules can absorb this excitation energy and emit new light called autofluorescence. Normally each absorbed photon results in a photon being emitted at longer wavelength (lower photon energy). This can be used to provide high resolution images of biological tissue and being "label-free", it avoids potential hazards associated with externally applied contrast agents. The properties of this light can provide information about the molecules that emitted it and this enables imaging of the biochemical or structural changes in tissue associated with disease, which can aid diagnosis. As well as the wavelength (colour) of the autofluorescence, the fluorescence lifetime (i.e. the rate at which the signal decays following excitation) can also provide useful "spectroscopic" information.

A key challenge for optical imaging in tissue, however, is optical scattering, which limits the depth at which optical images can be obtained. This can be addressed using the technique of multiphoton microscopy, which entails illuminating tissue with light at twice the wavelength (half the photon energy) that is usually absorbed. If the incident intensity is high enough, then the tissue can absorb two photons simultaneously to produce each autofluorescence photon. This permits excitation of tissue at longer wavelengths (which undergo less scattering and permit deeper imaging) and also provides 3D imaging. This is because the autofluorescence photons are only efficiently produced in the focal plane where the intensity is highest. Thus a multiphoton microscope provides "optical sectioned" images of slices of tissue that are in the focal plane. Currently, however, there is only one commercially available clinically approved multiphoton microscope, which is licensed for imaging skin, and no clinical multiphoton endoscope. We propose here to develop two new complementary approaches to in vivo clinical multiphoton imaging that should enable high resolution (subcellular) imaging of less accessible (including internal) tissues to provide real-time replacements for conventional histopathology - allowing in situ diagnosis and guiding of therapeutic interventions where it is critical to define lesion margins and minimise loss of function, e.g. in the brain, prostate, spine, skin (Moh's surgery) etc.

The first instrument is a novel lightweight hand-held multiphoton scanner, which would be able to compensate for patient motion (permitting longer acquisition times), richer spectroscopic readouts (including of fluorescence lifetime) and larger fields of view (to permit visualisation of whole lesions). This instrument would be applicable to image any external tissue or to tissues exposed during surgery.

The second instrument is a disruptive technology concept invented and pioneered at Imperial for ultracompact multiphoton endoscopes of unprecedented size (<400 microns diameter) and flexibility, for use via fine needles directly in the organ of interest or via thin anatomical channels (down to breast ducts). It could thus provide sub-cellular imaging almost anywhere inside the body, including under the guidance of other imaging modalities (e.g. ultrasound, MRI).

These instruments would be engineered for clinical trials following ex vivo human tissue studies during this project.

This project aims to provide new non-invasive tools for improved patient diagnosis and treatment and for improved capabilities for in vivo research. Thus this research should ultimately benefit patients and reduce costs to the NHS and other healthcare providers by providing faster optical readouts of tissue state ("optical biopsy"), by allowing a greater number of tissue sites to be probed in a clinical examination, by reducing the need for tissue biopsies and by enabling imaging of sensitive areas of tissue that cannot be biopsied. Examples include reducing or replacing the need for rapid cryo-histology during tissue-sparing surgery such as neurosurgery or Moh's surgery, thereby enabling these procedures to become faster and more accurate - hence improving patient outcomes. The proposed new instrumentation would also find other biomedical applications, e.g. for research, diagnosis and monitoring of therapies in cancer, inflammation, wound healing and any condition where label-free readouts of metabolism or tissue matrix properties are useful.

In vivo imaging technologies are also important for drug discovery and therefore the pharmaceutical industry, e.g. during preclinical drug testing in animals, and would permit longitudinal studies, therefore reducing the required numbers of animals and increasing the value of the data. Their application with fluorescence labels, including fluorescent proteins in genetically manipulated live disease models, could provide in vivo readouts of signal pathways for drug discovery and the label-free readouts based on autofluorescence could provide information on changes in metabolism and tissue matrix properties. Importantly, label-free readouts could be translated directly from animals to humans and used to evaluate new therapies, including drugs, and to monitor therapeutic procedures and outcomes in patients.

The most immediate impact of this project would be the subsequent medical research utilising the label-free detection, diagnosis and monitoring of disease afforded by the new instruments in the areas of our clinical advisory panel members, which include dermatology, gastrointestinal endoscopy, head and neck surgery and neurosurgery. We thus envisage patient-orientated follow-on studies and clinical trials to establish the patient benefit and/or cost-savings of these new clinical tools. For neurosurgery there is a specific unmet need for a minimally invasive functional and spatially-resolved readout of brain tissues that allows surgeons to decide whether it is necessary to perform further excisions and this would be one of our first targets.

Commercialisation would be an important route to realising the anticipated biomedical impact and this project would also stimulate the development of related instruments optimised for a range of applications, including with lower cost laser sources, thereby benefitting the instrumentation industry and the economy. We are seeking commercial exploitation partners and want to use the data from this project to build the prototypes needed to carry out clinical trials of the technology that are needed to validate the clinical applicability and strengthen our case for commercial investment. We already have strong links with biomedical instrument manufacturers including Mauna Kea Technologies (MKT, France) and JenLab (Germany) who are current market leaders in the field of in vivo clinical (endo)microscopy and note that the chief scientific officer of MKT will sit on our clinical advisory panel. MKT have already expressed interest the adaptive MP endoscopy technology, which could complement their existing clinically licensed endomicroscopes. Any exploitation of existing IP would be managed by Imperial's tech transfer company Imperial Innovations, in collaboration with Bath Ventures.

We also note the valuable training of project research staff who will gain interdisciplinary training at the interface between physics and medicine.