Getting a closer look at prostate cancer with molecular imaging

Consider this scenario: A man diagnosed with prostate cancer undergoes a radical prostatectomy. His prostate-specific antigen level, or PSA, drops from 12 ng/ml to nearly zero. But several months later, his PSA begins rising yet again, indicating that his cancer might be making a comeback.

What now? This patient’s CT and MRI scans show no evidence of cancer in other parts of the body, but they don’t show cancer in the hollow where the prostate once was, either. So where exactly is the cancer? Will salvage radiation therapy to the hollow, also called the prostate bed or prostatic fossa, drop the PSA to negligible levels permanently? Or will radiation do little more than cause troublesome side effects while the cancer continues its forward march elsewhere? Should the patient instead opt for a systemic treatment, such as hormone therapy, to control the cancer and lower PSA?

Answering these questions isn’t easy. Oncologists and urologists weigh many factors in making a decision about how to proceed, including how quickly the PSA is rising, the patient’s PSA and Gleason scores prior to surgery, and biopsy details — the location of the original tumor, for example. They plug this information into a formula or nomogram that predicts what’s likely to happen with the patient. But that’s all it is — a prediction. One study found that among 198 high-risk patients, clinical algorithms predicted only about 40% to 45% of instances in which disease spread to lymph nodes. So, aside from flipping a coin, how can one decide what to do?

The advent of molecular imaging

Aiming to help doctors locate metastatic cancers too small to be detected by CT or MRI in soft tissue, the FDA approved an agent called capromab pendetide (ProstaScint) in 1996. A monoclonal antibody, capromab pendetide “searches” the body for the one specific fixture it can lock onto: prostate-specific membrane antigen, or PSMA. All prostate cells express some PSMA, but cancer cells express much more of it, making them capromab pendetide’s primary target.

But scientists needed a way to see which cells capromab pendetide attached to. To solve that problem, they mixed it with minute amounts of the radioactive isotope indium-111. Like sailboats pulling up to a pier, sails raised, capromab pendetide docks with PSMA, waving its indium-111 flag. Devices called gamma cameras can then take pictures of the patient, showing doctors where indium-111 — and thus prostate cancer cells — have collected. Using this molecular imaging technique, doctors determine whether a local therapy for the prostate bed, such as radiation, would be best, or whether the patient should have a treatment to attack cancer throughout the body.

Since then, several studies have shown that indium-111 capromab pendetide scans can predict which patients will respond to salvage radiation therapy — that is, radiation therapy for prostate cancer after another treatment, such as surgery, has failed. (To read the complete reports of these studies, see “Early capromab pendetide studies: Pro,” below.) In a 1998 study, Iowa researchers looked at 32 men who had had a radical prostatectomy but whose PSA continued to rise following surgery. These men underwent an indium-111 capromab pendetide scan followed by radiation therapy to the pelvis. Seventy percent of the men with normal scans or scans showing cancer only in the prostate bed had a lasting reduction in PSA over the next 13 months, on average. In contrast, only 22% of the men whose scans showed cancer outside of the prostate bed had a lasting PSA response.

Researchers at the Lahey Clinic in Massachusetts came to similar conclusions. They performed indium-111 capromab pendetide scans in 48 patients with biochemical recurrence after radical prostatectomy; 13 subsequently opted for radiation therapy. Of that number, six had cancerous activity beyond the area that would receive radiation and seven did not. Radiation therapy failed in 66% of the patients with distant activity on the scans, but only failed in 29% of patients with localized cancer activity on the scan.

Despite the early positive press, indium-111 capromab pendetide scans never gained widespread acceptance. For one thing, the scans proved difficult to interpret, resulting in many false-positive readings; patients were diagnosed with metastatic cancer when they really didn’t have it. Shortcomings in imaging didn’t help: traditional methods lacked the resolution necessary to show critical anatomical details, so clinicians couldn’t reliably determine whether the indium-111 had accumulated in a lymph node or blood vessels. The scans’ utility was also limited by the fact that they couldn’t clearly show bone, which is the first site of metastatic prostate cancer in 72% of patients.

Subsequent studies questioned the value of the scans. (See “Later capromab pendetide studies: Con,” below). A 2003 study from the University of Cincinnati yielded results that were “discordant with prior reports.” Among 30 patients, the test failed to predict who would respond to radiation therapy and achieve biochemical control of cancer and who wouldn’t. In other words, men with a negative scan outside the pelvis didn’t fare any better after radiation therapy than those with a positive scan. “This surveillance method is of limited use,” the researchers concluded.

The following year, a team of Chicago researchers wrote in The Journal of Urology, “In its current form, this imaging technique is a considerable drain on time and financial resources. Our data do not justify its potential to replace or supplement traditional methods of assessing patients with biochemical failure after radical prostatectomy.”

Fusing images for greater precision

CT and MRI scans don’t exactly have a great track record of detecting metastatic prostate cancer in soft tissue either. Rises in PSA may precede clinical evidence of disease by a few years. And while larger lesions may be visible, small lesions may not. An early metastatic lesion usually measures less than 1 centimeter in diameter, which can be missed by CT and MRI. Furthermore, these techniques may indicate an enlarged node, but there are many causes of lymph node enlargement other than prostate cancer. Thus, a node that is enlarged due to a mild infection can be mistaken for metastatic disease.

Surgically removing the pelvic lymph nodes during a radical prostatectomy and analyzing them for cancer might seem like an obvious answer. But metastatic prostate cancer can spread to lymph nodes in the abdomen first, bypassing the pelvic lymph nodes. In fact, 15% to 50% of patients initially told that their pelvic lymph nodes don’t harbor cancer actually have prostate cancer in more remote nodes.

Technological advances in the last few years may renew interest in indium-111 capromab pendetide scans as a management tool for prostate cancer. Since the molecular scan gives information on function and the CT or MRI scan yields information on anatomy, fusing the images combines the data. (Image fusion has been done for years with different types of scans for this very reason.) So in theory, a metastatic lymph node seen on a lower resolution molecular scan can be pinpointed when it is laid on top of the higher resolution CT. In addition, the radiologist “reading” the images can better distinguish bowel and blood vessels from overlying lymph nodes, for example, resulting in fewer false positives and more accurate diagnoses.

A 2004 New York study argues the point. Researchers compared the utility of indium-111 capromab pendetide and MRI/CT scans alone and fused together in 58 patients with a rising PSA after primary therapy. Of the 161 sites considered positive for cancer according to the indium-111 technique, 74 were found to be negative after the images were fused with MRI/CT. An additional 13 sites not detected by indium-111 were found after the images were fused. The result: 25 patients had a diagnosis of metastatic disease changed to a diagnosis of an isolated, local recurrence of cancer. This same group of investigators found that the ability of indium-111 capromab pendetide to predict response to salvage radiation therapy increased to 69% when the images were fused with MRI/CT.

In addition, a May 2008 paper* published in The Journal of Urology favors the use of indium-111 capromab pendetide scans when fused with CT images. Investigators found that the technique could independently predict which patients diagnosed with localized prostate cancer had an increased risk of biochemical failure after brachytherapy, suggesting that the test could be used to evaluate patients for high-risk disease before treatment. (To read the full text of these papers and other recent studies on your own, see “Image fusion studies,” below.)

So does this mean that molecular scans with capromab pendetide should be done on every patient? No, at least not yet. The studies done thus far have included relatively few patients over a short period, and they’ve had mixed results. Even the manufacturer of the indium-111 agent notes in its literature that patient management should not be based on indium-111 scan results alone because of a high rate of false-positive and false-negative image interpretations.

In March 2008, Beth Israel Deaconess Medical Center, a teaching affiliate of Harvard Medical School, began offering the indium-111 capromab pendetide scan, making it just one of a few facilities in Massachusetts to offer the test. But demand has been limited, and only nine patients were scanned in the first three months. (See What’s involved in an indium-111 capromab pendetide scan?” below.) In not one of those cases did the test results alter the patient’s treatment plan or offer critical new information, so physicians there are still trying to determine its value. “It could go either way,” said one nuclear medicine physician.

Here’s the bottom line: the indium-111 capromab pendetide scan may be helpful in some subsets of patients, such as those with a biochemical recurrence following prostate surgery. Given the high rate of false positives and false negatives, results should be validated with other tests if there is a high degree of suspicion about the spread of cancer.

Originally published Oct. 1, 2008; last reviewed March 22,2011.