PET tests the waters of breast cancer monitoring

Researchers see tremendous potential but remain cautious because of limited reimbursement

By: Charles Bankhead

In several respects, the application of PET to breast cancer represents a success story in the development of the technology: reimbursable indications, potential for new indications, and continued evolution of the technology itself, most notably in the form of new tracers.

Despite that substantial upside, investigators and users remain cautious in their overall assessment of the role PET will ultimately play in diagnosis, evaluation, and management of breast cancer. Much of the caution revolves around the uncertain reimbursement climate.

On Oct. 1, PET received Centers for Medicare and Medicaid Services approval for reimbursement of selected breast cancer imaging studies, including studies related to monitoring treatment. The approval differs from those for other cancers, said Dr. Landis Griffeth, a radiologist at Baylor University Medical Center.

"With other tumors where CMS pays for PET imaging, the reimbursement is for initial diagnosis, initial staging, and treatment follow-up," Griffeth said. "The language is a little different for breast cancer. First of all, they will not pay for initial diagnosis. Also, it's not clear what is meant by 'treatment monitoring.'"

The problem lies in convincing payers to reimburse; many insurers have already announced they will not cover reimbursement for PET studies in breast cancer, he said.

As approved by CMS, PET can be used to evaluate a patient suspected of having recurrent breast cancer or distant metastatic disease. The indications have a potentially key role in therapeutic decisions.

"If a patient is scheduled to undergo mastectomy but there is a suspicion of metastatic disease, you would want to do a PET scan to determine whether the surgery is appropriate," Griffeth said. "The second arm of the approval relates to treatment follow-up. There is a fair amount in the literature to say that PET will be useful after the first few courses of chemotherapy to determine whether a therapeutic regimen is going to work."

The area generating most enthusiasm for PET imaging in breast cancer relates to monitoring therapy. Researchers and clinicians uniformly express optimism about the imaging modality's potential role in therapeutic decision-making.

"If the uptake intensity of a lesion on PET decreases after one or two courses of chemotherapy, that has a very high predictive value for patient response," Griffeth said. "On the other hand, if the intensity of the lesion does not decrease after the first few courses of therapy, that probably means the therapy is not working and the physician should consider switching to a different type of therapy."

A European study demonstrated that FDG-PET imaging of primary and metastatic breast cancer after an initial dose of chemotherapy might predict the likelihood of pathologic response (J Clin Oncol 2000;18:1676-1688).

"The data suggest that we can begin to see evidence of a response after a single cycle of therapy and begin to make decisions very early about who is going to respond and who is not," said Dr. David Mankoff, an associate professor of radiology in the nuclear medicine division at the University of Washington in Seattle. "While the results are not perfect, they do suggest that functional imaging techniques can provide an early assessment of response."

Similar data have continued to emerge over the past two years, indicating that PET likely has an important role in monitoring breast cancer therapy. Collectively, the data show that PET is the most reliable examination for evaluation of response to

therapy, according to Dr. Farrokh Dehdashti, an associate professor of radiology at Washington University in St. Louis.

PET's role in monitoring therapy includes predicting the outcome of therapy. Investigators at the University of California, Los Angeles reported that FDG-PET proved statistically superior to conventional imaging modalities for predicting the risk of disease progression or death after completion of breast cancer therapy (J Nuc Med 2002;43:325-329).

"Metabolic changes occur before anatomic changes. PET can detect these changes before they can be seen with other types of imaging," Griffeth said.

TAILORING THERAPY

Investigators continue to push PET's capabilities for predicting response to breast cancer therapy. Mankoff and his associates reported that initial FDG uptake by a breast tumor and the ratio of FDG uptake to blood flow, as assessed by O-15 water, might identify patients whose tumors are likely to respond

to cytotoxic therapy (J Nucl Med 2002;43:500-509). The results showed that high levels of glucose uptake predicted poor response to neoadjuvant chemotherapy, whereas a low FDG metabolic rate relative to blood flow predicted a complete response.

Such findings hint at the possibility of using PET to tailor breast cancer therapy to the tumor characteristics of individual patients. However, the PET application in breast cancer becomes more complicated at that point and might require tracers other than FDG, Mankoff said.

The ideal situation would be to identify tumor characteristics that predict which tumors will respond and which won't before initiating therapy. The best option for that application is fluorine-18 fluoroestradiol (FES), an estrogen analog that binds to estrogen receptors in ER-positive breast cancer, Mankoff said. This tracer has been studied extensively by a group at Washington University.

Imagers might also be able to predict which patients will be resistant to therapy. If a PET study using FES finds that the patient has lost her estrogen receptors, a nonhormonal therapy will be necessary. Imaging early in the therapeutic process should give an indication of changes in tumor biochemistry. Most of the work to date has been done with FDG, but there are theoretical and practical indications that other tracers under development might do even better.

Among potential FDG alternatives, FES has one of the longest histories. Development of radiolabeled estradiol dates back to the 1980s, Griffeth said. In vitro and in vivo data have demonstrated its ability to predict response to tamoxifen therapy.

"Several tracers are more specific than FDG for evaluation of breast cancer," Dehdashti said. "The most promising is FES, which can be used to predict response to hormone therapy. We have published several papers describing the use of this tracer in breast cancer, but this is purely research."

Another tracer that has shown promise is fluorine-labeled thymidine (FLT), which homes in on proliferating cells. Investigators at UCLA have begun evaluating the tracer in the hope that it will prove to be a good marker for response to chemotherapy, said Dr. Johannes Czernin, director of nuclear medicine in the department of medical and molecular pharmacology. His group has also done preliminary studies to explore the potential clinical utility of radiolabeled estrogen receptor ligands and labeled tamoxifen.

Because FLT can be used to evaluate cell proliferation, the tracer could have potentially widespread application in assessment of response to therapy, not only in breast cancer but in other cancers as well, said Dr. Homer A. Macapinlac, director of the PET section at the University of Texas M.D. Anderson Cancer Center.

"A lot of treatments we have for cancer involve the use of antiproliferative agents," he said. "If we can perform an imaging test to identify the volume of tumor that is still actively multiplying before treatment and then repeat the scan after the patient has started treatment, we might be able to identify very early on which patients will respond to treatment."

PET might also have a significant role in monitoring response to therapy in the subset of breast cancer patients who have bone-dominant metastatic disease. Although such patients rarely achieve cures, their disease can be responsive to various forms of therapy, and a patient might live 10 to 20 years after diagnosis of the stage IV disease, Mankoff said. The current approach to monitoring therapy in patients with bone-dominant breast cancer is to use a combination of bone scans and MRI, but this has proved suboptimal.

"Bone scans and MRI are not very good for following response to therapy," he said. "The bone lesions don't change much or might even change in the wrong direction when the patient responds. Medical oncologists who treat breast cancer would tell you that one of the biggest problems they have involves assessment of bone lesions. Physicians have trouble telling whether any progress is being made."

In a recently published study, Mankoff and his Seattle colleagues reported that whole-body FDG-PET might be useful for following response to treatment in patients with bone-dominant breast cancer. Their study reports that results of serial PET scans correlated significantly with clinical assessment of response to treatment and with changes in the values of a particular tumor marker (Acad Radiol 2002;9:913-921).

"Better approaches might come along in the future, but right now a simple FDG-PET using a semiquantitative approach with a standard uptake value seems to be a reasonable way to monitor response to therapy in breast cancer patients who have bone-dominant disease," he said.

DIFFERENCES OF OPINION

Opinions differ with respect to potential expansion of indications for PET imaging in breast cancer. According to UCLA's Czernin, PET could potentially play a major role in evaluating a variety of patients who prove problematic for mammography and other imaging modalities. These include women who have dense breast tissue, scarring in breast tissue, and implants and those who have dense breast tissue and are at increased risk for breast cancer for any reason. Czernin acknowledges, however, that PET's role in breast cancer ultimately hinges on the reimbursement rules.

"I hope people will show enough reason to allow us to use PET to its fullest capabilities in breast cancer," he said. "Radiologists acknowledge there are shortcomings with mammography; that is why they do more MRI and ultrasound studies and work on cancer protocols to improve diagnostic sensitivity. There is no question that PET should be part of that process."

Other PET users and investigators are less optimistic, particularly with respect to the likelihood the imaging technology will have a role in the initial diagnosis of breast cancer.

"If PET were allowed to be used for initial diagnosis, it would have a significant role, particularly in patients who have dense breasts," Griffeth said. "At this point, we won't be allowed to use PET that way. In reality, PET is more accurate than any other modality for assessing whether a patient has distant metastatic disease and for assessing local recurrence."

A battery of tests, including CT, bone scans, and often MRI for local assessment, are currently standard procedure at some centers. Based on available data, PET has proved more accurate than any of those modalities, due to its ability to look at metabolism instead of anatomy, Griffeth said.

Conceding that his views don't win much support in the nuclear medicine community, Mankoff said that he does not expect PET to play a huge role in the diagnosis of breast cancer.

"My concern is that nothing has shown any particular advantage for being able to image the biology of very early breast cancer," he said. "The earliest forms of breast cancer-low grade and in situ-tend to have low FDG uptake. Because of that I have been pessimistic that FDG-PET or any of other nuclear medicine tracers on the horizon will play a big role in early diagnosis of breast cancer, with the exception of specific niches."

MR looks more promising for secondary diagnosis or high-risk screening than nuclear medicine procedures, according to Mankoff.

"That doesn't diminish the excitement over the nuclear medicine piece. It's just that I don't see right now that we have that magic bullet that makes it simple to diagnose early breast cancer," he said.

Dehdashti concurs that PET will not have a role in breast cancer screening.

"The technology cannot compete with mammography or MRI for detection of primary breast cancer, due to its limited resolution. So PET is not going to be used for breast cancer screening in the near future," he said.

Although PET is not very good for assessment of regional lymph node involvement, Dehdashit said, it might be useful for assessment of internal mammary or supraclavicular lymph nodes, which are difficult to evaluate by physical examination or anatomical imaging. PET does have a role in evaluation of regional lymph node involvement in patients who have locally advanced breast cancer treated by neoadjuvant chemotherapy prior to surgery, and it is helpful in subsets of patients with breasts that are difficult to evaluate mammographically. Neither indication qualifies for reimbursement, however.

PET's ultimate contribution to diagnosis and evaluation of breast cancer will likely be determined by clinical experience, Griffeth said.

"Only after a technique is in relatively widespread clinical use do we get the data we need to tell us how to optimally utilize the test," he said. "Only after widespread clinical use will we know in which groups of patients PET is likely to make a significant contribution to care."

Despite having created a substantial role in oncology imaging, PET remains underutilized, according to Czernin.

"We have not even scratched the surface of opportunities. Sometimes it is mind-boggling just how little it is used," he said.

Czernin's hope is that PET's superior capability for distinguishing malignant and benign masses eventually will win over the support necessary to expand its use.

Mr. Bankhead is a freelance writer in Houston.