Moritz F. Kircher, MD, PhD

Enlarge Image Raman image-guided brain tumor resection Cartoon based on actual data using SERS nanoparticles in a mouse glioblastoma model. SERS nanoparticles were injected intravenously, and Raman imaging was performed 24 hours later. Green = brain tumor; gray = healthy brain; red = Raman signal. One of the most important goals in oncology is the earliest possible detection of a cancer, at a stage where the tumor can still be completely removed or destroyed – either by surgical excision or minimally invasive interventional radiology techniques.

Such a goal can only be achieved if the tumor burden is assessed with the highest possible sensitivity and accuracy, and then removed completely. This requires imaging methods that allow sensitive whole-body tumor detection and ultra-high-precision intraoperative tumor delineation, including detection of tiny micro-metastases or fingerlike tumor projections that currently cannot be visualized with existing methods.

In Vivo Raman Imaging of Nanoprobes

One new method that promises to allow such ultra-high-sensitivity detection is surface-enhanced Raman spectroscopy (SERS) imaging. SERS imaging is fundamentally different from other known imaging methods. Raman imaging detects unique photons that have interacted with particular atomic bonds in molecules. Because the interaction of a Raman photon with a particular atomic bond of a molecule causes a specific change in the wavelength of the Raman photon, the molecule can be detected with unequivocal precision: Raman allows “molecular fingerprinting.

Enlarge Image Principle of Raman Effect and SERS Imaging (A) Raman effect: Incident photons (λi) are mostly scattered elastically (Rayleigh scattering), with scattered wavelength remaining the same (λs=λi). A few photons are scattered inelastically (Raman scattering), which increases their wavelength (λs>λi). Raman spectra depict intensity vs. energy difference between incident and scattered photons. (B) Molecular imaging agent approach showing SERS nanoparticles, which consist of a metallic core, a unique Raman active layer adsorbed onto the metal surface, and a shell coating the nanoparticle. An array of unique spectral signatures can be achieved by modifying the Raman active layer on each nanoparticle. (C) Intrinsic approach showing a piece of human tissue being irradiated with a laser source. Intrinsic Raman spectral signatures of tissue can reveal important information about the molecular composition of cells or tissue of interest.

This principle can be used to detect targeted molecular nanoprobes that carry a unique spectral fingerprint, which has several major advantages over existing imaging methods:

• The Raman fingerprint unequivocally identifies the presence of the contrast agent (no issues with autofluorescence).
• Using sophisticated nanotechnology, it is possible to enhance the signal intensity of the Raman nanoprobe many orders of magnitude, allowing markedly higher sensitivity than other imaging methods. Enlarge Image Raman nanostars
• Raman active nanoprobes do not exhibit photobleaching (in contrast to fluorescent probes, for example).
• Raman fingerprints are unique. This allows creation of large libraries of nanoparticles with different targets or properties that can all be imaged simultaneously (“Raman multiplexing”).
• Raman nanoprobes can be based on nontoxic, inert materials (gold).

Imaging based on SERS nanoprobes therefore provides the highest possible sensitivity and signal specificity as well as multiplexing capabilities, a combination of attributes that can pave the way for cancer detection and destruction with unparalleled precision in intraoperative and endoscopic applications.

Multimodal Nanoprobes for Cancer Imaging

The goal to allow both pre- and intraoperative tumor detection requires that the Raman nanoprobes also be “seen” with a whole-body, three-dimensional imaging modality, such as MRI, PET, or CT. Another more recently developed three-dimensional modality is photoacoustic imaging, which can allow for deeper tissue localization in the operating room. We are able to add these additional modalities to the nanoprobes by incorporating respective contrast agent molecules into the nanoparticle formulations. This results in multimodal nanoparticles that can be localized by three or more modalities, each of which has complementary strengths (Nature Medicine, in press). We are collaborating on some of these projects with Sam Gambhir of Stanford University.

Multimodal Theranostic Nanoprobes

Ideally, cancer cells could be selectively destroyed during imaging. This goal can be achieved by adding therapeutic capabilities to diagnostic nanoprobes that selectively accumulate in cancer cells (theranostic nanoprobes). For example, nanoparticles can be tuned to generate heat while being imaged, and as a result the locally generated heat could potentially destroy small foci of cancer cells that cannot be removed by resection.

About Moritz Kircher

Moritz Kircher received his medical and doctorate degree from the Humboldt University in Berlin, Germany. He did a  radiology residency and fellowship training in MRI at Harvard and Stanford Universities and postdoctoral training at the Center for Molecular Imaging Research of Harvard Medical School/Massachusetts General Hospital and in the Molecular Imaging Program at Stanford. In 2010, he joined the faculty of Memorial Sloan-Kettering Cancer Center as a physician-scientist, focusing on novel molecular imaging approaches for early cancer detection. He is a member of the Department of Radiology, the Nanotechnology Center, and the Brain Tumor Center.