Imaging Cancer Gene Activity in Patients From Outside the Body

Researchers in Philadelphia inject a nuclear medicine gene probe into patients, then use scanning techniques to see if active cancer genes light up.

Abstract

Is the lump malignant or benign? The investigators propose to answer this question by injecting a nuclear medicine gene probe into patients with suspicious lumps in their breasts. They then use scanning techniques to see if active cancer genes light up.

The American Cancer Society estimates that more than 500,000 U.S. cancer patients will die in 2006 within 5 years of the date on which they were first diagnosed with cancer. When such a catastrophic disease as cancer is suspected, a most urgent question arises: Is the lump malignant or benign?

All diseases result from activation of genes — those of the patient; those of a virus, bacterium, fungus, or parasite; or both. The early detection of cancer gene activity deep within the human body might answer the most urgent question rapidly, enabling life-saving early intervention.

DETECTION OF CANCER

Diagnostic imaging of abnormalities began with X-rays and was improved by computerized tomography (CT). These tools have been joined by high-frequency sound scans (ultrasonography), nuclear medicine scans — including single photon emission computed tomography (SPECT) and positron emission tomography (PET) of radioisotope decay — and magnetic resonance imaging (MRI) of protons in water.

In the case of breast cancer, self-palpation and X-ray mammography are advocated and commonly practiced for early detection. Nevertheless, these methods miss 20 to 40 percent of lesions, particularly in the dense breasts of younger women.

When an abnormality is found, an interventional procedure must be performed to obtain a tissue sample from the malignant lesion for histologic examination. Approximately 60 to 80 percent of histologic examinations reveal that these abnormalities are benign. In cancerous cells, the characteristic alterations in cell morphology occur much later than the genomic mutations that lead to the initial uncontrolled cell proliferation and to malignancy.

Therefore, in principle, targeting the characteristic fingerprints of malignant transformation could detect cancer at an earlier stage than conventional histologic examinations. Early detection of cancer can warrant aggressive therapeutic intervention — surgical, chemical, or radiological.

FINGERPRINTING CANCER

Fingerprinting disease processes and cancer markers provides the next step in functional imaging. The chelator sestamibi, radiolabeled with technetium (Tc)-99m, can reveal some breast tumors by SPECT imaging. Glucose, which is concentrated by the most active tissues, can be labeled with the radioisotope fluorine-18 to enable detection of tumors, inflammation, or highly active brain zones by PET imaging. SPECT imaging with synthetic peptides, such as octreotide labeled with indium-111 and antibodies labeled with Tc-99m, now reveal specific receptors and other proteins on cells that indicate cancer and other diseases.

These imaging methods are valuable but not efficacious in detecting all cancers. For example, these modalities miss up to 30 percent or more of breast cancers. None can be used successfully to detect other cancers, such as prostate cancer.

CANCER GENES

In the case of cancer, we must consider thousands of genes that are involved in uncontrolled cell proliferation. In cancerous cells, some of the normal cell-growth genes sustain an activating mutation. Although there are hundreds of different types of cancers and genetic analyses of tumors have led to an estimated 200 different genes that are mutated in various cancers (Wooster 2000), the typical tumor displays only five or so mutated genes (Hilgers 1999), as Albert Knudson of Philadelphia’s Fox Chase Cancer Center predicted 36 years ago (Knudson 1970). Furthermore, each gene can be mutated at different locations within itself. These characteristics make each cancer unique, and every tumor is a mixture of subtly different cancer cells.

CANCER GENE IMAGING

Abundant recent research in the literature reports that a large array of characteristic fingerprint proteins encoded by oncogenes are overexpressed, exogenously on the cell surface or endogenously within the cytoplasm. If one could image active cancer genes from outside the body, a precise gene-targeted intervention could be directed at the onset of disease.

Such imaging methods are under development. One can hypothesize that external imaging of excess cancer gene activity in radiographically suspicious masses will correlate with malignancy, eliminate unnecessary surgical interventions, and detect cancers at a very early stage. In our translational research during the past 6 years, we have designed, synthesized, characterized, and radiolabeled certain biomolecules that specifically targeted those oncogenes in human tumors that were growing in immunocompromised mice.

We focus our efforts on early detection of breast, pancreas, or colorectal cancers by targeting oncogenes CCND1, KRAS, MYC, and VPAC1. To detect the oncogene mRNAs, we use radiolabeled peptides or peptide nucleic acids for planar, SPECT, or PET imaging. We chose Tc-99m for planar or SPECT imaging, or Cu-64 for PET imaging.

Most breast cancer cells over-express the oncogene VPAC1 cell surface receptor, which binds a peptide ligand called vasoactive intestinal factor (VIP). Following positive preclinical imaging results, we initiated a feasibility study using a Tc-99m-VIP analog for imaging tumors in humans (Thakur 2000). Negative controls did not display appropriate concentration of labeled VIP. Out of 11 patients examined thus far, there was concordance in 9. In the other 2 patients, only Tc-99m-VIP was positive for tumors known to express VIP receptors. One resulted from recurrence of resected breast cancer (Figure 1) and the other from a recurrence of neurofibroma in the neck.

FIGURE 1.

Scintigraphic imaging of breast tumors with Tc-99m-VIP. A 42-year-old woman with prior left mastectomy presented with recurrence in right breast and left operative site. Lateral images with Tc-99m-sestamibi (left) show uptake in the chest wall and right breast (arrows). Left-side view (center) obtained 15 minutes after injection of Tc-99m-VIP and right-side view (right) 1 hour after injection of Tc-99m-VIP show same lesions (arrows), perhaps with better intensity than on corresponding Tc-99m-sestamibi images (left).

SOURCE: THAKUR 2000, REPRINTED WITH PERMISSION OF THE SOCIETY OF NUCLEAR MEDICINE

Clinical and experimental data strongly suggest that high levels of cyclin D1 protein, encoded by the CCND1 gene, as well as other oncogenes, are implicated in breast cancer development. The insulin-like growth factor 1/insulin-like growth factor 1 receptor (IGF1/IGF1R) system plays a major regulatory role in development, cell-cycle progression, and the early phase of tumorigenicity (Baserga 1995). The IGF1R gene is amplified in about 70 percent of human tumors, particularly in metastatic cells (Armengol 2000).

The strategy for imaging elevated CCND1 mRNA depends on three elements (Figure 2). First, we use peptide nucleic acid as a rugged and specific hybridization probe. Second, we add a chelator that can hold a radionuclide tightly. Third, we add a peptide analog that will bind to IGF1R and allow entry of the radioprobe into the cancer cells.

FIGURE 2.

Tc-99m-chelator-PNA-peptide, WT4185, designed to bind to the cell surface IGF1 receptor, internalize, and hybridize with CCND1 oncogene mRNA. Top: Ball-and-stick model. Bottom: Molecular scheme. Scintigraphic imaging of gamma particles emitted upon decay of Tc-99m identifies sites of high oncogene CCND1 mRNA expression.

PNA=peptide nucleic acid.

We have been able to validate our hypothesis by noninvasive SPECT imaging of CCND1 mRNA at 6,000 copies/cell in breast cancer xenografts in immunocompromised mice by specific hybridization of Tc-99m-WT4185 (Figure 3), while control sequences showed only faint signals in tumors (Tian 2004). We also have imaged MYC mRNA breast cancer xenografts (Tian 2005) and KRAS mRNA in pancreas cancer xenografts (Chakrabarti 2005) with Tc99m-peptide nucleic acid (PNA)-peptide probes with appropriate sequences. PET imaging of cancer gene expression is also under development, using Cu-64 as a source of positrons (Chakrabarti 2005).

FIGURE 3.

Scintigraphic images of gamma particles emitted by decaying Tc-99m in nude mice carrying human MCF7:IGF1R estrogen-receptor-positive breast tumor cell xenografts (arrowhead). Images were obtained 12 hours after injection of PNA-free control (WT990), PNA mismatch control (WT4172), peptide mismatch control (WT4113), and CCND1 PNA antisense probe (WT4185).

SOURCE: TIAN 2004, REPRINTED WITH PERMISSION OF THE SOCIETY OF NUCLEAR MEDICINE

In the future, external detection of which cancer genes are active or not active in breast lumps might improve early treatment of breast cancer and ultimately reduce deaths and suffering from breast cancer.

THIRD-PARTY PAYERS

According to current procedural terminology (CPT) codes, SPECT reimbursement can be as low as $300, or as much as $1,000, depending on the carrier. Similarly, PET reimbursement can be as low as $800, ranging up to $3,000, again depending on the carrier. The actual costs and the level of reimbursement probably will not change on introduction of mRNA imaging probes. Although expensive to produce under current good manufacturing practice protocols, the mRNA imaging probes will require only 10 micrograms for each patient administration, with radiolabeling on site according to standard procedures.

Once sufficient clinical trial data exist to merit approval of a particular mRNA imaging probe by the U.S. Food and Drug Administration, obtaining reimbursement for the existing CPT codes might not garner significant resistance from payers.

FUTURE DIRECTIONS

In addition to determining the nature and gene activity of a suspicious lump, radioimaging of mRNAs might be able to identify early cancers that have not formed detectable masses. The simplest example is ductal carcinoma in situ (DCIS), where there are clearly some cancerous cells in the nipple aspirate but current methods cannot determine where the cancer is located. Early localization of the cancerous zone would allow a less-disfiguring option than total mastectomy. The cancer gene-oriented approach also might prove valuable in early detection of ovarian, pancreas, colon, and prostate cancers.

In the future, it might be possible to image cell-surface disease markers with near infrared fluorescence (NIR). Investigators are pursuing the NIR option with fluorescently labeled antibodies and with fluorescent quantum-dot nanoparticles conjugated with antibodies. Diffraction of fluorescent light limits NIR detection to tissues that are at most a few centimeters deep.

SPECT and PET imaging currently lead this translational research direction, but other modalities such as MRI or NIR might follow soon. One can readily imagine using external mRNA imaging for genetic diagnoses of other diseases, such as heart, joint, or brain inflammation.