Optical Spectroscopic Detection and Imaging of Breast Cancer

Institution: Lawrence Livermore National Laboratory
Investigator(s): Stavros Demos, Ph.D. - Rajen Ramsamooj, M.D. - Stavros Demos, Ph.D. -
Award Cycle: 2001 (Cycle VII) Grant #: 7DB-0091 Award: $0
Award Type: TRC Pilot Award
Research Priorities
Detection, Prognosis and Treatment>Imaging, Biomarkers, and Molecular Pathology: improving detection and diagnosis

This is a collaboration with: 7DB-1091 -

Initial Award Abstract (2001)
The development of mammography and the implementation of regular screening programs have had a significant impact on breast cancer mortality. However, there remain the problems of both sensitivity and 'false negative' mammograms, which result in unnecessary follow-up and biopsies. We are examining an alternative approach to breast cancer detection using optical methods combined with laser-based technology. Some key benefits of using light for medical imaging are the use of non-ionizing radiation and the ability to monitor small, localized regions either externally to the body or internally through small optical fibers. We predict that research in this area will provide both detection advantages and, in addition, give information regarding tissue structure, status and composition.

This project is a collaborative effort to develop optical cancer detection and imaging technology with special attention to diagnosing breast cancer. The presence of characteristic optical "signatures" of tissue components are key to the success of any photonic imaging modality for biomedical applications. These optical "signatures" may arise from differences in the biochemical and/or structural characteristics of the tissues. The differences at the cellular level (size of the cells and the nucleus, density and makeup) between cancerous and normal tissue offers the basis for the development of such optical imaging technology and novel instrumentation for quick, accurate, and minimally invasive cancer detection. This technology has the potential to provide detection and imaging of cancerous lesions 1 mm in diameter or smaller, which is more than an order of magnitude better than mammography.

To achieve our goals we will explore two independent approaches. The first approach takes advantage of the polarization property of light. When polarized light is propagating in a tissue, the polarization of the scattered light is sensitive to the structure of the tissue. This property can be used to separate out different tissue components, such as those present in cancerous cells and tissues. Additional information can be obtained by performing a multicolor analysis of this effect. The second approach utilizes the properties of tissue biomolecules to absorb light and re-emit at a different spectral region. We predict that this re-emission effect will reveal cancer specific biomolecules that differentiate the cancer from normal breast tissue. Preliminary experimental results show the presence of such emission signatures from breast cancer under 'photoexcitation' with red laser light. Thus, we plan to expand these observations and explore the sensitivity and selectivity limits of both the polarization and re-emission properties inherent in our optical/laser methods. In all these experiments, we will use fresh surgical samples where the diagnosis of the lesion is. We will also be able to perform independent biochemical analysis to validate the optical/laser results.

The long term objective of our research is to develop: a) minimally invasive detection techniques for cancer using optical fibers inserted in very thin needles, and b) cancer imaging systems to be used during lumpectomy to assist the surgeon detect and excise cancer. We will also investigate the cancer specific optical "signatures" using microscopic techniques to reveal their origin at the cellular level.

Final Report (2004)
Note: This grant was extended 6-months to complete the funding and aims. The following abstract was presented at the 24th IABCR meeting in Sacramento in November 2003.

Breast cancer remains a devastating disease affecting as many as 1 in 8 women. Breast excisions are one of the most common surgical operations in the United States. Although the preoperative imaging of breast cancers may help stratify the risk that a breast abnormality is malignant, excisional biopsies of the breast are still performed for purely diagnostic purposes. The mammographic and ultrasound criteria for malignancy are well validated, though there still remains a large fraction of lesions with indeterminate preoperative imaging that mandate histologic assessment to exclude malignancy. Previous studies have shown that different types of tissue have unique native emission spectra when subjected to laser irradiation. We are developing a laser-based spectroscopic imaging system which is able to digitally capture the spectral images of tissues with resolution spanning from the tissue level to the single cell level. We hypothesized that this technology can offer the surgeon several advantages that would have a clear impact on multiple levels. Should optical imaging become validated for breast cancer, the immediate information obtained includes: 1) the presence or absence of a malignancy in the lumpectomy specimen, 2) the histological type of malignancy (in situ vs. invasive), and 3) the status of the margins if the lumpectomy is performed for malignancy.

Fresh breast tissues containing cancer were examined by a pathologist immediately after excision. Thin sections of tissue (2-3 mm) containing cancer and normal breast tissue were examined by optical spectroscopy methods using two imaging systems under development. The first system utilizes laser excitation at 405, 532 and 632 nm in order to record low resolution, auto-fluorescent images in the far- and near-infrared spectral region. In addition, polarized light scattering images are recorded with illumination at 700, 850 and 1000 nm. The second system is a prototype hyperspectral microscope that offers spatial resolution of 1 ?m to record spectroscopic images of tissue microstructures and individual superficial cells. The imaging techniques are similar to the ones described above for the first system with the addition of spectroscopic examination extends in the near ultraviolet and visible part of the spectrum. The examined tissues were processed in the usual fashion for paraffin sectioning. The resulting optical spectroscopic images were compared to the histologic section of the same tissue. The experimental results demonstrate an increased fluorescence from malignant tissue compared to the normal breast. This difference can be used to distinguish various tissue components. The hyperspectral microscope investigation also shows that cancer margins can be detected with single cell spatial resolution.

We have developed spectroscopic imaging methods that can take advantage of differences in the optical characteristics between breast cancer and normal breast tissue, in vitro. This new technology can potentially allow the real time detection of breast cancer in a minimally invasive manner. While optical imaging would not replace existing modalities, it would supplement the analysis by providing an immediate result for the surgeon. Therefore, the surgeon could combine the diagnostic and therapeutic aspects of the operation to reduce or eliminate the need for repeat operations as well as provide information on the status of the surgical margins. This technology can help reduce morbidity and mortality as well as increase survival.