Smart probes and biomarkers spot earliest signs of cancer

November 1, 2008

Molecular imaging is rapidly advancing as a biomedical modality that increases the understanding of underlying cellular mechanics and dynamics and adds a new dimension to the diagnosis and treatment of disease. It may be a sensitive and specific method for evaluating cancer by measuring the expression of genes that trigger oncogenesis and stratifying tumors on the basis of their biological characteristics.

Molecular imaging is rapidly advancing as a biomedical modality that increases the understanding of underlying cellular mechanics and dynamics and adds a new dimension to the diagnosis and treatment of disease. It may be a sensitive and specific method for evaluating cancer by measuring the expression of genes that trigger oncogenesis and stratifying tumors on the basis of their biological characteristics.

Central to any advancements is the development of biomarkers and targeted smart probes and the equipment to image them.

In the Cassen Lecture given at the 2008 SNM meeting, Dr. Mathew L. Thakur shared some of his thoughts about the profound influence that genomic biomarkers will have on molecular imaging and clinical practice in the future.

Thakur is director of the laboratories of radiopharmaceutical research and molecular imaging and a professor of radiology, radiation oncology, and nuclear medicine at Thomas Jefferson University in Philadelphia.

The following remarks have been excerpted from that presentation as well as a follow-up interview with Diagnostic Imaging.

Cancer remains the most formidable disease of humanity. In 2007, 7.6 million people died of cancer throughout the world, and 11 million more were treated for it. At the same time in the U.S., more than 565,000 people died because of cancer: more than one a minute every day and every night.

Cancer remains a disease of the cell. Life begins as a single cell, which multiplies into an estimated 75 trillion cells by adulthood. In order to sustain life, cells go through enormously complex chemical processes all their lives and all the time.

Researchers believe that modulation of the chemical network leading to damage of DNA transforms cellular chemistry into a chaotic process that creates many different biomolecules and pathways into the nucleus and the cytoplasm. These molecules, which are both endogenous and exogenous, show up long before morphologic changes appear on imaging studies or pathologic changes appear on histology, the gold standard for the analysis of cancer.

So it is a simple concept that if we target these molecules, we should be able to shed light on the pathogenesis of cancer and develop biologic tools that may be used to create new tools for detecting and managing cancer. Investigators in academia and industry are engaged in oncology research that will translate this knowledge into practical applications not only to detect cancer at an early stage but also to stratify it nonsurgically, treat it, monitor the effectiveness of the treatment, and extend the quality of life of those affected by it.

Managing the disease nevertheless requires a better description of the genetic damage that drives human cancer and better interpretation of the complex network of proteins that are needed for cells to survive.

Great progress has been made in the treatment of cancer: External-beam radiation, the use of chemical agents, and removal of tumors with surgical excision remain the mainstays of cancer treat-ment. But these modalities also kill normal cells along with malignant ones. Future drugs, however, will target the molecular and genetic products that foster cancer growth. What will matter most in the future will be the molecular characteristics of cancer.


One part of the emerging molecular imaging story involves the targeting of receptors, such as VPAC1, that are overexpressed on the surface of cancer cells. I will share examples from my laboratory in which we studied the use of a technetium-99m-labeled peptide that binds specifically and avidly to VPAC1.

A 22-year-old woman with neuroblastoma had redness and tenderness on the left side of her neck, which could have been due to metastasis, particularly in bone. Bone and sestamibi scans were negative. Within 15 minutes of injection of the Tc-99mlabeled peptide specific for VPAC1 receptors, however, a 3-mm spot was found. The lesion was excised, and histological analysis concluded it was high-grade spindle cell sarcoma.

In four other patients who were suspected of having breast cancer, Tc-99m sestamibi scans were negative. After the administration of the Tc-99m-labeled peptide, however, abnormal lesions were revealed by an extensive uptake. Histology as well as reverse transcriptase polymerase chain reaction confirmed that VPAC1 receptors were expressed on the cells of the lesions.

Although these are small indications, the examples demonstrate that a specific molecule may have abilities to detect cancer that cannot be found with current imaging modalities. The other part of the molecular imaging story involves the targeting of specific oncogenes. We have turned our attention to oncogene peptide nucleic acid (PNA) probes that may diagnose cancer with scintigraphy, MRI, or optical imaging technologies.

One of these probes, labeled with positron-emitting copper-64, binds to oncogenes such as CCND1, which is overexpressed in breast cancer cells, and K-RAS, which is overexpressed in pancreatic cancer.

In experiments in nude mice bearing human breast tumor xenografts, the Cu-64 peptide PNA uptake was significantly greater four hours after injection than it was in xenografts that were scanned with a mismatched PNA probe, which served as a control.

Uptake of the Cu-64 peptide was low in normal tissues at both four and 24 hours after injection. PET/CT imaging of Cu-64 in mice bearing breast tumor xenografts showed even more intensive uptake. The Cu-64 probe targeted the CCND1 gene in animals that bore xenografts of estrogen receptor-positive human breast tumors.

In pancreatic cancer, we found a similar situation. There was increased uptake of the Cu-64 probe in transgenic mice with pancreatic cancer, which is known to express the K-RAS gene.

At present, if we want to confirm the presence of any malignancy, we have to take a biopsy into histology, and the pathologist distinguishes cells that are normal from those that are malignant. But morphology changes much later than the biochemistry of the cell.

Since biochemical modulation initiates and drives the cancer development process, it’s possible that molecular imaging may be able to detect cancer without waiting until morphological changes have occurred and without having to take biopsy samples from tumors.


We are studying VPAC1 receptors in prostate cancer because detecting the disease at an early stage in men is just as difficult as detecting early-stage breast cancer in women. Of 750,000 prostate biopsies done every year in the U.S., two-thirds find benign pathology.

We have been using the Cu-64 probe to study a group of transgenic mice. These mice spontaneously grow prostate cancer by two to eight months of age and express VPAC1 receptors on the tumors. In one example, we found differences in two sibling mice. One had a large uptake of Cu-64; the other did not. We sacrificed the mice and could visibly see prostate cancer in the mouse that had exhibited intense Cu-64 uptake. Histologic analysis of the prostate tissue showed grade 4 hyperplasia.

The clinical literature indicates that 85% of men who have grade 4 hyperplasia develop prostate cancer within six months, and the remaining 15% develop cancer within nine months or a year. If we can detect hyperplasia with a Cu-64 VPAC1-specific peptide probe, we will be able to detect prostate cancer before it can be found with current imaging techniques. We would minimize unnecessary biopsies and be able to start early treatment.


The role of the Ki-67 protein was determined only five or six years ago, yet already there is a microchip that quantifies its expression. The Ki-67 protein is located in the nucleus of the cell, and its expression is strongly associated with cell proliferation. The protein is, in fact, absent in cells at rest. As a result, the expression of the Ki-67 protein in excised tissue samples has become the hallmark of cancer diagnosis and prognosis.

Use of the Ki-67 protein as an in vivo molecular imaging tool is not yet sought out, however, because it is used in conjunction with serial biop-sies. There is no reason why we cannot target Ki-67 protein expression noninvasively to signal the presence of cancer.


If a microscope allows us to see a single cell ex vivo, imaging equipment should enable us to detect small clusters of cancer cells in vivo. Researchers at Washington University, Stanford University, and other sites are developing photoacoustic and Raman spectroscopic devices that can image structures deep into the body with high resolution and sensitivity.

One of these techniques uses nanotubes that are bound to antibodies specific for receptors expressed by one or more types of cancer. When the nanotubes are exposed to cancer cells, they produce an electric pulse that indicates there is a cancerous cell.

This phenomenon will allow investigators to construct a microchip-type device that may be digitized and converted into an oncometer that we will be able to carry around in our pockets. Such an instrument could identify that cancer cells are circulating in the body but have not yet formed into clusters. In the future, I believe, a self-operated oncometer will indicate whether a patient has a susceptibility for cancer.

Another technique that fuses ultrasound and laser light in photoacoustic imaging has been used to acquire 3D images of not only subcutaneous melanomas but the blood vessels that feed them.

Quantum dots are a new class of contrast agent for multimodal imaging. The fluorescent dots are nanocrystals of 1 to 10 nm in size that are brighter than conventional contrast media, so only a small number of dots is needed to produce a signal. The dots are also stable, so images may be acquired over a long period of time without deterioration of image quality. Quantum dots may be used in a variety of ways to diagnose, stage, and manage the cancer patient.


A look into the future sees many exciting possibilities. Nuclear medicine and molecular imaging physicians will be able to determine which gene and how many copies of it are expressed in a cancer cell; image malignant cell clusters in microns, not millimeters or centimeters, with high sensitivity and specificity; stratify cancer noninvasively; and assure that “bad” genes will not be humankind’s destiny. Visionary scientists predict that medicine will change more in the next 20 years than it has in the last 2000. Molecular imaging will play a major role in those changes. The possibilities are endless. They are limited only by imagination.