APRIL 2003

Molecular imaging finds role in therapy assessment

Nuclear medicine creates gateway to first human applications

By: James Brice

Guided by their traditional orientation toward disease as a physiological process, nuclear medicine physicians are routinely using molecular imaging in clinical practice, while researchers are developing new techniques that use radiotracers to expand its role.

Nuclear imaging researchers do not see molecular imaging as the next step for medical imaging; as far they are concerned, it is already a reality. They argue that the actual genesis occurred in the mid-1990s with the launch of indium-111 OctreoScan, the first FDA-approved peptide-based radiopharmaceutical agent.

There is no denying that OctreoScan was targeted to exploit specific protein behavior. The agent's designers used an eight-peptide residue that binds to somatostatin subtype-2 peptides. Somatostatin is overexpressed in neuroendocrine tumors, making OctreoScan an attractive choice for imaging that type of cancer.

Fluorine-18 fluorodeoxyglucose (FDG), which is by far the most popular PET tracer, is also a molecular agent, said Dr. Sam Gambhir, the director of the Crump Molecular Imaging Institute at the University of California, Los Angeles. FDG is a glucose analog that is transported into cells on the gluc-1 transporter. Gluc-1 is present in many kinds of cells but is wildly overexpressed in cancers and is responsible for the relatively high accumulation of F-18 FDG in metastatic cells. After entering a cell, FDG is phosphorylated by hexokinase, and the change in electrical charge traps the FDG inside the cell.

FDG-PET illustrates why nuclear imaging is well suited for molecular applications, said Dr. Markus Schwaiger, director of research at the Institute for Radiology, Technical University of Munich in Germany. Nuclear imaging is incredibly sensitive; it can detect picomolar concentrations of substances in tissue. A picomole is one-trillionth or 10-12 of a mole. By comparison, MRI is sensitive to the 10-4 molar level, meaning that PET can produce quantifiable results based on far less molecular activity.

Other advantages include a large library of radiolabeling techniques to assist drug design and the availability of microPET and microSPECT scanners that aid the transition from animal testing to human research. Disadvantages are the short half-life of some radioisotopes, low spatial resolution, the high cost of instrumentation, especially when an on-site cyclotron is required, and past problems gaining FDA approval and Medicare reimbursement for PET radiopharmaceuticals.

"Taken together, however, nuclear imaging is in my opinion the most mature technology to target molecular structures," Schwaiger said.

Researchers are optimistic about expanding the role of FDG-PET to therapy monitoring. It will challenge CT, the current choice for monitoring the effects of chemotherapy, Gambhir said. Oncologists rely on CT measures of tumor diameter and volume to determine whether chemotherapy is working. But it can take many months before changes are seen, and at that point, the disease may have progressed too far for a change in therapy to be of any benefit.

In contrast, measurable changes in FDG uptake rates are observed much sooner, according to Dr. Lale Kostakoglu, an associate professor of radiology at New York Presbyterian Hospital.

"Early prediction means you can change treatment if the tumor is refractory to treatment," she said.

Sufficient evidence already exists to show that FDG-PET can reliably measure the therapeutic response to chemotherapy administered for advanced lymphomas and lung cancer, Kostakoglu said. A study comparing FDG-PET and CT after lymphoma therapy found relapse among all patients with a positive post-therapy PET study, but only a 25% relapse rate for the patients with positive CT. The positive predictive value for PET was 100%, compared with 42% for CT (Blood 1999;94:429-433).

Experience with FDG-PET in predicting the response of non-small cell lung cancer to chemotherapy has also been encouraging. A Duke University study of 113 patients found that an FDG-PET scan after first-line treatment provides a good indication of the patient's prospects for survival. The mean survival time for patients with a positive scan was 12 months, while 85% of those with a negative scan were still living 38 months after the procedure (AJR 2000;173:769-774).

It is less certain whether FDG-PET can serve as a surrogate marker for radiotherapy. Cancer cells that show signs of radiation-induced inflammation can metabolize FDG before they die, in a manner that could be confused with uptake associated with active cell growth, Kostakoglu said.

MEASURING CELL PROLIFERATION

F-18 fluorodeoxythymidine (FLT) is considered another good candidate for a surrogate marker of therapeutic response. FLT is a thymidine derivative that is phosphorylated at a rate equivalent to DNA production, said Dr. Anthony F. Shields, a professor of oncology at the Karmanos Cancer Institute in Detroit. DNA production slows as cell division slows.

"It is one of the earliest responses to chemotherapy," Shields said.

FLT could possibly work in tandem with FDG. Although the two agents travel different molecular pathways, both are taken up and trapped in cells in a similar manner.

"You may say that imaging glucose use is not that interesting, but it has become a valued clinical tool," Shields said. "FLT is just measuring a different pathway."

FLT potentially offers a faster measure of response than FDG because DNA synthesis associated with cell division turns off faster than glucose utilization. FLT signal is also confounded by glucose uptake associated with inflammatory processes. Because brain cells rarely divide, FLT is an attractive alternative.

According to Schwaiger, FLT could outperform FDG for measuring the response to cytostatic chemotherapies designed specifically to inhibit cell proliferation. Although FLT has yet to be tested in large-scale trials, several small studies indicate the agent's potential. University of Washington researchers uncovered an extremely strong correlation (p>0.0001) between the maximum pixel standard uptake value (maxSUV) of FLT and Ki-67, an immunohistochemical measure of cell proliferation, in a study of 10 lung cancer patients (Clin Cancer Res 2000; 8[11]:3315-3323).

IMAGING MULTIDRUG RESISTANCE

Technetium-99m sestamibi, which has gained broad acceptance as an alternative to thallium-201 for myocardial imaging, may also be used someday to identify the early signs of multidrug resistance to chemotherapy in cancer patients. Multidrug resistance is a galling phenomenon: While cancers can develop defenses against a specific chemotherapy, some cancers can actually become immune to an entire class of cancer-killing drugs.

Multidrug resistance is associated with the overexpression of one or more of at least 10 different molecules. One mechanism is encoded by the MDR1 gene and its product, P-glycoprotein, said Dr. David Piwnica-Worms, director of the molecular imaging center at Washington University in St. Louis. P-glycoprotein is a transmembrane drug efflux pump; when overexpressed, it pumps chemotherapy out of cancer cells faster than the agent can enter them. As a result, the therapy never has a chance to work.

As they studied how Tc-99m sestamibi works in the heart, Washington University researchers discovered that the agents could potentially measure multidrug resistance. They learned that the lipophilic cations of sestamibi are rapidly trapped in the mitochondria in the heart muscle. Because heart myocardial cells have no P-glycoprotein and lots of mitochondria, the agent is readily taken up by those cells, Piwnica-Worms said. That biochemical behavior makes sestamibi an excellent myocardial imaging agent.

In contrast, P-glycoprotein resides in abundance on the excretory surfaces of bile cells, and as a result, sestamibi is rapidly excreted from the liver. The rapid washout of Tc-99m sestamibi can thus be used to mark the presence of P-glycoprotein and to predict the patient's susceptibility to multidrug resistance. Conversely, high uptake and retention following administration of Tariquidar (XR9576) or Amdray (PSC833) may indicate that these investigational agents have successfully inhibited P-glycoprotein, removing it as an obstacle to successful administration of conventional chemotherapies.

DETECTING APOPTOSIS

The ability to use a cell's natural mechanism to induce programmed death is considered a great opportunity for new genetic therapy. Scientists have discovered the roles of specific genes and associated proteins in recognizing when a mutation happens during cell division. Mutations are fairly common, but they rarely cause damage because mutated cells are programmed to kill themselves before dividing again.

Scientists have also learned how this programming is lost during oncogenesis. They are devising strategies that would reintroduce the capability of cancer cells to induce their programmed cell death. In this context, apoptosis markers such as annexin-V could be extremely valuable, according to Gambhir.

"There are many instances where we should be concerned about apoptosis. If you implant cells in the heart as a means of repair, you want to know if those cells are dying. And during chemotherapy, you want to know if the treatment is triggering cell death," he said.

Moreover, annexin-V could potentially play a useful role in other clinical situations where cell death should be measured. It might help assess myocardial infarction, differentiate recurrent tumor from necrosis, aid in the assessment of atherosclerotic plaque stability, and monitor tumor response to therapy, according to Dr. H. William Strauss, clinical chief of nuclear medicine at Cornell University.

Annexin-V binds to membrane-associated phosphatidylserine, an enzyme that is pumped out of cell membranes during the early stages of programmed cell death, Strauss said. Annexin-V can be labeled with F-18 for PET imaging, gallium-67 or Tc-99m for SPECT imaging, and fluorescine or rhodamine for optical imaging.

Strauss is working with technetium-labeled annexin-V to examine atherosclerotic plaques for signs of apoptosis.

"We can examine various targets that identify the status of the atheroma. One way is to examine the inflammation itself with FDG-PET because the inflammatory cells themselves consume glucose. These cells undergo apoptosis after they have tried to clean up the lesion. The marker for such programmed cell death can be seen with annexin-V labeled with Tc-99m or iodine-124," he said.

Other researchers are examining the potential of annexin-V for assessing myocardial apoptosis. Dr. Neil Steinmetz, medical director of Theseus, the developer of annexin-V, and colleagues at eight medical centers reported that Tc-99m annexin-V imaging identified persistent myocardial tissue injury in 48 of 59 patients who had experienced a myocardial infarction in the past 96 hours. The study was presented at the 2002 Society of Nuclear Medicine annual meeting. In another presentation at the same meeting, Dr. Raymond Taillefer from the Hospital of the University of Montreal provided evidence that the procedure is effective during the first four days following an infarction.

Phase II clinical trials suggest that annexin-V may measure the response to therapy for lymphoma, sarcoma, and breast and lung cancers. An animal study found that response to the initial dose of chemotherapy for hepatoma was measured with Tc-99m annexin-V just six hours after treatment (J Nucl Med 2003 Jan; 44[1]:92-97).

Overall, the molecular imaging environment is extremely fluid. Continuous progress is being made on these and more strategies that researchers believe will establish the foundation of future molecular imaging practice. In nuclear medicine, molecular techniques such as FDG-PET have already made an impact on cancer diagnosis and staging. Some protocols for assessing the response to therapy are ready for routine use.

And this is only the beginning, according to Schwaiger. He expects that the sophistication of nuclear imaging applications for the assessment of cancer treatment will increase, as will their use, as new molecularly based therapies are introduced. In the future, therapies will be tailored to the biology of individual patients and diseases.

"That is why we are so interested in applying molecular imaging for the in vivo characterization of tumor biology," he said.