WASHINGTON, DC—Two new protein-based tests may provide much-needed improvement over the standard prostate-specific antigen (PSA) test, according to data presented here at the American Association for Cancer Research Annual Meeting.
ProPSA May Distinguish Aggressive Cancers
Scientists at Johns Hopkins University recently reported that proPSA, an inactive but circulating form of the PSA enzyme, is a more accurate marker of cancer than are the total, free, or complex forms of PSA, which are used in the standard PSA test (Sokoll et al: Urology 2003; 61:274–276).
In that study, the group tested archival serum samples from 119 men with a total PSA of 2.5 to 4 ng/mL and found that the ratio of proPSA to total PSA more accurately identified the cancer patients in the group.
Even in this small sample, use of the proPSA ratio would have significantly decreased the number of unnecessary biopsies, according to the authors.
Now, a new study shows that the predictive value of proPSA holds up when tested in a larger population, said one of the coauthors of the initial study, Stephen D. Mikolajczyk, of the Molecular Development Department at Beckman Coulter, Inc.
In this larger retrospective study, he and his colleagues measured the percentage of proPSA relative to total PSA in serum samples from 1,091 men enrolled in cancer screening studies and compared it with the percentage of free PSA in the same samples.
Of those men, 555 had a total PSA between 2 and 4 ng/mL, 320 of which had biopsy-confirmed benign disease and 235 had cancer.
With the percent proPSA test, the investigators were able to accurately detect 90% of the cancers, and would have eliminated the need for 19% of the biopsies in this group. This is a statistically significant improvement over the percent free PSA test in this low total PSA range.
The percent free PSA is not giving you any better specificity than “flipping a coin” in the 2 to 4 ng/mL range, said Mr. Mikolajczyk, whereas the proPSA is.
“More important, proPSA is significantly elevated in the samples with a Gleason score above 7. That tells you that proPSA is associated with a preferential elevation in aggressive cancers.”
Specifically, the proPSA levels are about 50% higher in patients with a cancer that scores a seven on the Gleason scale, than it is in men with cancers that score six or below. Therefore, not only would a proPSA test be able to detect cancers earlier than the currently available tests, but it may also be able to help physicians and patients decide whether to take a watch-and-wait approach or a more aggressive therapeutic one.
The proPSA test is an antibody-based test, similar to what is used for standard PSA tests in clinics now. Mr. Mikolajczyk predicts that the test could be available in as little as two to three years.
Finding Prostate Cancer's Fingerprint
Using a completely different approach, researchers at the NCI and the FDA have used proteomics to try to improve the existing PSA test.
With proteomics scientists are no longer restricted to looking for an elusive unique protein that signals a cancerous growth. Instead they can look at a whole spectrum of proteins from sera or other tissues to find a pattern that differs from that of healthy patients.
“This is a new era for us, pattern diagnostics. It is something that is being driven by the gene microarray field right now,” said Emanuel Petricoin, PhD, Co-Director of the FDA's Laboratory of Immunology in the Center for Biologics and Research, a lead author of the study.
Already researchers use gene expression patterns from microarrays to distinguish diseased tissue from healthy controls and different cancer types.
“ProPSA is significantly elevated in the samples with a Gleason score above 7. That tells you that proPSA is associated with a preferential elevation in aggressive cancers.”
More recently, Dr. Petricoin and his collaborator, Lance Liotta, MD, PhD, Chief of the Laboratory of Pathology in the Center for Cancer Research at the NCI, have developed methods to diagnose cancers based on the mass spectrometry profile of serum proteins and computer algorithms. [Watch for an upcoming “OT Primer” column for a description of how this method works.]
However, because PSA can already detect cancers, the team's goal in this case was to provide a second level of non-invasive testing that could refine the results obtained from a standard PSA test. Specifically, they wanted to find a pattern or fingerprint from serum proteins that would tell doctors which of the patients who have a PSA score in the grey zone between 2.5 and 10 ng/mL have cancer and which have benign disease. Such a test would immediately reduce the need for biopsies in a large fraction of patients, said Dr. Petricoin.
To identify such a prostate cancer-specific pattern, the team first analyzed serum samples from 63 men with PSA levels between 2.5 and 15 ng/mL, 33 of whom had had at least one negative biopsy and 30 with biopsy-confirmed cancer.
Using these protein profiles they trained the computer algorithm to see a difference between a cancer specific pattern and a benign pattern. They then tested it on 91 blinded samples. The protein pattern test successfully distinguished all 28 of cancer cases from the 63 men who had negative biopsies.
“We had an accuracy of 97% for proven prostate cancer, even in the indeterminate range,” said Dr. Petricoin. In this small sample alone, the researchers estimate that 42 of the 63 negative biopsies (67%) could have been avoided if this test was used in addition to the standard PSA test.
The team is currently working to confirm these results on a larger population. They plan to develop a similar test that can distinguish indolent cancers from life-threatening ones, but that project is only in the very early stages.
FDG-PET Seen as Key Prognostic Tool for Metastatic Thyroid Cancer
FDG-positron-emission tomography (PET) appears to be a valuable tool in evaluating prognosis and guiding management of patients with metastatic thyroid cancer, according to a retrospective study of 400 patients presented at the Society of Nuclear Medicine Annual Meeting.
Presenting the findings, Qiang Wan, MD, a resident at Memorial Sloan-Kettering Cancer Center, said the scans would be particularly useful in the commonly occurring case of metastatic thyroid cancer patients who are thyroglobulin positive, but iodine negative.
Dr. Wan reported that a negative PET scan indicated the cancer was slow growing. That's relatively good news for the patients, he said, meaning they could expect to survive another five to six years, and would guide surgeons to restrict any surgery to low-risk procedures.
However, a positive PET scan indicated a more aggressive cancer, with a patient's risk of death rising by eight-fold compared with the PET-negative patients, Dr. Wan said. That would signal the need for more aggressive management.
In the study, with a median follow-up period of 22 months since the first PET scan and longest follow-up at 72 months, 70 people died, he and his colleagues reported.
“Those with negative PET scans had dramatically lower Tg [thyroglobulin] levels than those with positive PET scans. Furthermore, PET scans can often disclose residual disease even in low-risk—Stage I-III—patients who appear to be disease free.”
Some distant metastases were detected only by PET scan, resulting in a change in therapy. As a single diagnostic test, PET demonstrated distant metastases in 16.5% of patients with clinical Stage I-III disease.
All patients in the study had histologically documented thyroid cancer, following total thyroidectomy. The decision to obtain a PET scan was based most commonly on an elevated Tg with a negative RAI scan.
A positive PET scan reflected a focal lesion with a standardized uptake value (SUV) higher than 2. Ditant metastases were defined as those outside the neck and upper mediastinum.
Co-principal investigator Steven Larson, MD, Chief of the Nuclear Medicine Service at Sloan-Kettering, said that up to half the patients will get into that situation. He described the group of metastatic thyroid cancer patients who are thyroglobulin positive but become iodine negative as “fairly important.”
Dr. Wan said he previously had reported on 100 patients, in which he indicated the emerging value of FDG-PET for thyroid cancer. But this latest report based on 400 patients represents the first presentation of a survival curve.
In development is a formula to include age and thyroglobulin, iodine, and PET status that predicts how long the patient will survive, Dr. Wan added.
Sloan-Kettering has scanned about 1,000 thyroid cancer patients with PET, although most are lost to follow-up, he noted. The cancer center has about 10 new thyroid patients a week.