Targeted PET agents fulfill valuable diagnostic roles

May 1, 2005

Radiopharmaceutical developers are eternally optimistic about the potential of molecularly targeted probes to determine the presence, extent, and response of cancers to therapy. The researchers have been adept at identifying promising agents that perform well during in vitro tests and small-animal experiments, but they have not been particularly successful at producing clinically viable products.

Radiopharmaceutical developers are eternally optimistic about the potential of molecularly targeted probes to determine the presence, extent, and response of cancers to therapy. The researchers have been adept at identifying promising agents that perform well during in vitro tests and small-animal experiments, but they have not been particularly successful at producing clinically viable products.

Except for the triumph of fluorine-18 fluorodeoxyglucose (FDG) and the lesser success of technetium-99m sestamibi, no targeted radiopharmaceutical agent developed in the last 10 years has had a major impact on patient diagnosis or treatment.

That situation is destined to change soon. New agents are moving through the research and development pipeline, and proliferation agents, probes capable of mapping hypoxia, hormone receptor agents, and reporter gene imaging protocols are making the transition from animal studies to early trials with human subjects.

"We are beginning to enter a new era where molecular imaging will become a clinical reality," said Dr. Steven Larson, director of nuclear medicine at Memorial Sloan-Kettering Cancer Institute. "It is not yet in full-scale clinical practice, but I would expect that is going to occur in the next two to three years."


Researchers described their progress toward that goal in a plenary session at the 2004 RSNA meeting in Chicago.

Touted as the greatest innovation since Tc-99m sestamibi, F-18 FDG has established the clinical foundation for PET. FDG is a chemical analog of glucose that measures cellular metabolism. It is profoundly sensitive to cancer because metastatic disease metabolizes FDG at 10 or more times the rate of surrounding tissue, except in the brain and in the presence of inflammation. As a result, cancer lights up like a firefly. Despite FDG's diagnostic powers, however, oncologists would like to find agents that more quickly detect the response of cancers to therapy, and they would like an alternative for detecting metastases and primary cancers in the brain.

That explains the considerable attention being paid to Dr. Anthony Shields' work with PET agents capable of imaging cellular proliferation. An agent that noninvasively reports the rate of cell division could complement FDG and its focus on metabolism. A reduction in cell division might be an early sign that cancer therapy is working. Brain cells divide slowly, if they divide at all, creating an opportunity for high contrast between background noise and signal originating from brain neoplasms.

Several studies examining the value of diagnostic imaging after radio- or chemotherapy have shown that F-18 FDG-PET imaging can take up to three months to reveal the effects of therapy, according to Shields, an oncologist and associate director for clinical research at the Barbara Ann Karmanos Cancer Institute in Detroit.

Early work with carbon-11 thymidine demonstrated that it is taken up by proliferating cells and is retained in the DNA synthetic pathway. Thymidine detects a slowing of cell division that precedes cell death promoted by cancer therapy, Shields said. But the half-life of C-11 is too short and thymidine degrades too quickly to permit its adoption as a clinical agent.

Thymidine's attributes, however, were preserved with the synthesis of F-18-3'-fluoro-3' deoxythymidine (FLT), a probe developed in Shield's lab, and F-18 2'-fluoro-5-methyl-1-Beta-D-arabinofuranosyluracil (FMAU), a compound developed by Dr. Peter Conti and his colleagues at the University of Southern California. Only about 2% of FLT, an analog of thymidine, actually gets to the DNA. FMAU does not degrade like thymidine, but it shares its ability to be phosphorylated by thymidine kinases and incorporated into DNA, according to Shields.

Progress has been made in simplifying the synthesis of FLT and in investigating its potential as a cancer imaging agent. John R. Grierson, Ph.D., a radiology research scientist at the University of Washington, developed a precursor for its radiosynthesis in 2000. GE Healthcare now sells FLT synthesis boxes. In February, the FDA approved an investigational new drug application (IND) that the National Cancer Institute had filed in 2004.

A large multicenter clinical trial of FLT is pending that will involve the University of Washington, Johns Hopkins University, Virginia Commonwealth University Medical Center, and Massachusetts General Hospital. Dr. John Hoffman, chief of the National Cancer Institute's molecular imaging program, is the principal investigator.

Encouraging results with F-18 FLT in animal trials have been reported when the probe was used to monitor therapy. Hamamatzu University researchers in Japan, for example, found that F-18 FLT uptake in tumors in mice fell significantly only six hours after a single session of radiotherapy. Residual effects remained measurable three days later (JNM 2004;45[10]:1754-1758).

At the RSNA conference, Shields displayed serial FLT images demonstrating the different patterns that appeared in breast cancer patients who responded or failed to respond to neoadjuvant therapy before sur-gery (Figure 1). For responders, the amount of uptake recorded at an average of 22 days after therapy was substantially lower than the pretreatment pattern. Normal uptake in the sternum and bone marrow, a proliferative tissue, was observed. Follow-up FLT-PET revealed that the benefit persisted 113 days after therapy. For the nonresponders, the uptake pattern changed little when measured about three weeks after therapy.

In the brain, FLT has shown promise for imaging high-grade primary tumors, Shields said (Figure 2). The difference between a mass and normal tissue background is 0.2 to 0.3 standard uptake values. The agent is poorly suited for imaging low-grade tumors that exhibit low levels of cell division, however. A separate study of 17 patients, conducted by Dr. David C.P. Cobben at Groningen University Hospital in Groningen, the Netherlands, indicates that FLT is inferior to F-18 FDG for staging and restaging non-small cell lung cancer (JNM 2004;45[10]:1677-1682).


In the first human study of F-18 FMAU conducted by Shield's group, dynamic PET imaging with the agent clearly visualized active tumors in the breast, brain, lung, and prostate (Eur J Nucl Med Mol Imaging 2005; 32[1]:15-22). The probe clears rapidly from the blood and concentrates in the liver. Unlike FLT, FMAU also accumulates in the heart because of the presence of thymidine kinase II. For reasons still not fully understood, normal bone marrow does not take up FMAU as aggressively as it takes up FLT, giving FMAU a possible edge for bone metastasis imaging, Shields said.

An early example of FMAU imaging in a breast cancer patient (Figure 3) confirmed the presence of a primary tumor in the right breast as well as axillary lymph node involvement. But the agent is considered a poor candidate in the upper abdomen because of robust liver and heart uptake, he said.

Hypoxia poses a huge problem for effective cancer treatment. It can immunize cancer cells against radiotherapy and alter gene expression to increase the aggressiveness of cancers and their resistance to chemotherapy.

Eppendorf O2 electrode histography is the gold standard for plotting tumor hypoxia, but the technique is invasive, and the results can be inconsistent. Over a decade ago, University of Washington researchers synthesized F-18 fluoromisonidazole (F-MISO) as a potential alternative, said Michael Welch, Ph.D., cochief of radiological sciences at the Mallinckrodt Institute of Radiology. Misonidazole undergoes bioreduction in hypoxic cells, where it is retained while being washed out of nonhypoxic cells. When F-18 F-MISO was tested on non-small cell lung cancer patients, however, the Seattle group found no correlation between hypoxic volumes and responses to therapy.

These limitations were at least partially addressed with Cu(II)-diacetyl-bis(N4-methylthiosemicarbazone) (ATSM), a probe developed by Welch and Yasuhisa Fujibayashi, Ph.D., a professor of molecular imaging at Fukui Medical University in Japan. ATSM can be labeled with various copper-based radionuclides, including Cu-60, Cu-61, and Cu-64. By altering the mitochondrial voltage potential of ATSM, the researchers found that it was trapped only in hypoxic tissues. The uptake pattern appears about 15 minutes after intravenous administration.

In vitro studies indicate that much more ATSM than F-MISO is retained per unit of oxygen in a given tissue volume, Welch said. Additionally, an in vivo study of cervical cancer patients established a correlation between the tumor-to-muscle ratio of ATSM uptake and hypoxia-related expression of VEGF, EGFR, and 1a COX-2 genes. (Eur J Nucl Med Mol Imaging 2003;30[6]:844-850). These findings led Welch to predict that radiologists will eventually hit the diagnostic equivalent of a home run with ATSM.

"We can predict response to therapy. We can predict disease-free survival, and we've shown that Cu-60 ATSM may be a way to direct radiation therapy," he said.


Welch has also been on the forefront of research examining the potential of zero-to-zero hormone receptor imaging. Working with John Katzenellenbogen, Ph.D., a professor of chemistry at the University of Illinois, Welch and his University of Washington collaborators have labeled ligands for the estrogen, androgen, and progesterone receptors.

Experiments with estrogen receptor labeling led to the synthesis of 16a-[18F]fluoro-estradiol (FES), a PET agent that has shown promise for predicting the response of ER+ metastatic breast cancer to tamoxifen therapy. At the 2003 Society of Nuclear Medicine meeting, Dr. David A. Mankoff, director of nuclear medicine at the University of Washington, described results using F-18 FES in 34 women with recurrent metastatic breast cancer. All 10 responders to hormonal therapy expressed high FES uptake.

Welch's experience with the agent has also shown a strong relationship between FES uptake and response to tamoxifen therapy. SUVs of at least 20 produced positive and negative predictive values of 79% and 88%, respectively, for whether the patient would respond, he said.

Successful androgen-receptor labeling led to F-18 fluoro-5a-dihydrotestosterone (FDHT), a promising agent for therapy measurement for men with metastatic prostate cancer and rising PSA levels. In a study of 20 subjects, all 12 who demonstrated FDHT uptake in PET imaging performed before flutamide (an androgen-receptor antagonist) therapy responded to the treatment, Welch said.

A study of 15 prostate cancer patients conducted by Washington University researcher Dr. Farrokh Dehdashti established the ability of F-18 FDHT to characterize the extent of metastatic disease. Her paper, presented at the 2003 SNM meeting, reported abnormal FDHT uptake in 10 patients and revealed multiple bone lesions and lymph node metastases that were confirmed with CT imaging.

Larson and his group at Memorial Sloan-Kettering are also involved in this line of research. Bench experiments found that FDHT is transported into cells where it binds to an androgen receptor before it is transported into the nucleus, he said. Larson established the feasibility of in vivo targeting and F-18 FDHT in a 2003 study reported at the SNM meeting (JNM 2004;45[3]:366-376). Among seven prostate cancer patients, he found that tumor uptake was rapid and retention was prolonged (Figure 4).


Although many diagnostic imaging strategies have a molecular basis, reporter gene imaging actually enables radiologists to render opinions about the diagnostic implications of subcellular biochemical events.

The main role of reporter gene imaging is to provide information about gene expression, especially the changes caused by therapies designed to alter the genetic makeup of diseased cells, said Dr. Sanjiv (Sam) Gambhir, director of molecular imaging research at Stanford University. Reporter genes can indicate whether the altered genes have penetrated their cellular targets and whether those new genes have been expressed. Cells can also be permanently tagged with a reporter gene in order to track its survival and migration in a living subject.

"Right now, gene therapy investigators are shooting blindly," Gambhir said. "When they deliver genes into a tumor or the heart, they have no way to follow what happens with the expression of those genes or the movement of those cells."

This imaging technique actually involves two major components: a reporter gene that is introduced into the nucleus of a cell and a reporter probe that carries a radioisotope. A gene delivery vehicle, often a sterilized adenovirus, transports the reporter gene to its nuclear target. A biochemical switch, called a promoter, drives transcription of the therapeutic gene and induces the production of a specific enzyme.

The reporter probe readily washes in and out of unaltered cells, assuring low background signal, but it is phosphorylated by the enzyme and trapped in cells where the reporter gene resides. The buildup of the tracer inside the nucleus produces a signal correlating with sites of therapeutic gene expression.

More than 400 published papers have established the efficacy of the PET reporter gene/probe approach, Gambhir said. Other notable contributors include Ronald Blasberg, Ph.D., at Sloan-Kettering, Dr. Peter Conti at the University of Southern California, Dr. Juri G. Gelovani at MD Anderson Cancer Center, Harvey R. Herschman, Ph.D., at UCLA, Dr. David Piwnica-Worms at Washington University, and Kurt Zinn, Ph.D., at the University of Alabama.

Gambhir secured an investigational new drug (IND) application for the reporter probe F-18 FHGB pencyclovir in October 2004. Nine patients with hepatocellular cancer had been examined with a reporter gene imaging technique using the pencyclovir probe when Gambhir spoke at the RSNA conference in December. In one case, viral particles carrying an altered gene that induces programmed cell death were injected at two sites in the tumor, using ultrasound guidance.

To detect gene expression, the fluorinated pencyclovir derivative was then administered systemically. PET imaging performed an hour later detected two points of gene expression corresponding with the injection sites and the phosphorylating effect of the reporter gene.

Reporter gene imaging has produced its biggest return thus far by monitoring the effects of gene therapy over time, Gambhir said. To illustrate those capabilities, he showed PET images acquired two and nine days following initiation of gene therapy. After the second day, considerable gene expression was reported from the targeted tumor, but after nine days, little activity was seen.

"This told us that the immune system had recognized the presence of a foreign gene and had shut it down," he said. "This shows us that we need to re-prime this patient with the gene by delivering more virus."

The potential clinical applications of reporter gene imaging are vast, according to Gambhir. His Stanford University team is developing cell trafficking techniques to label stem cells before implantation in infarcted myocardium. Immune cells used in cancer trials can be tagged and followed as well.


The group is also developing fusion reporter genes capable of producing proteins that fluoresce when they encounter a specific substrate. Such fusion genes create opportunities for multimodality imaging with PET and bioluminescent optical imaging to report molecular events, Gambhir said. Their applications are primarily in cancer evaluation but also extend to the characterization of cardiovascular disease and neurodegenerative disorders.

"Now that we have an IND, it is easy for different centers to use these techniques with the FDA's approval to perform human trials," Gambhir said.