PSA:Assay Change Makes 3.0 the New 4.0 for PSA Cutoff by WHO Standard

ASCO GU: Assay Change Makes 3.0 the New 4.0 for PSA Cutoff

By Crystal Phend, Senior Staff Writer, MedPage Today

Reviewed by Zalman S. Agus, MD; Emeritus Professor University of Pennsylvania School of Medicine.

A quiet switch in laboratory standards has artificially lowered prostate specific antigen (PSA) levels reported to physicians, confusing thresholds for biopsy and potentially leading to missed cancers, researchers reported here.

Assays calibrated to the WHO standards gradually adopted since 2000 yield PSA levels 22% to 25% lower than those under the old Hybritech standards, Amit Gupta, MD, MPH, of Memorial-Sloan Kettering Cancer Center in New York City, warned at the Genitourinary Cancers Symposium.

"This, of course, has a big impact on how we interpret PSA, which is so basic to urology," Gupta told MedPage Today.

When his group analyzed nationally-representative data, they found that close to 34% of cancers would be missed if the traditional PSA cutoff of 4.0 ng/mL -- determined under the old standards -- continued to be used with assay values determined under the new standards This scenario is all too likely, Gupta cautioned. Hybritech was the first PSA assay manufacturer on the scene in the 1980s, he explained, and other manufacturers followed suit in calibrating to this industry standard.

By the '90s, though, differences in antibodies and design between assays led to large variations in results, which WHO aimed to resolve with a new reference for calibration. Not all manufacturers have switched to the new standard yet, and when they do switch, laboratories may notify physicians only in the fine print of reports, if at all, Gupta said. Most lab reports do not indicate standard was used, he added, and only one company has a physician education campaign regarding the change. "The entire literature is based on the Hybritech standard, but more than 50% to 60% of PSA assays around the country are now calibrated to the WHO standard," he said in an interview. Few urologists and even fewer primary care physicians are aware of the issue, agreed Nicholas J. Vogelzang, MD, medical director of the Developmental Therapeutics Committee of US Oncology, who wasn't involved in the study.

Community oncologist Peter Lindberg, MD, of Northern New Mexico Cancer Center in Los Alamos, said that was true for him. "I thought that if you stuck with the same laboratory you wouldn't have a problem," Lindberg said.

This may be an issue that guidelines need to get involved in, to clarify that approximately 3.0 ng/mL has to replace the 4.0 ng/mL PSA threshold everyone is used to thinking of if used with the WHO standardized assays, Vogelzang said.

Vogelzang said he only stumbled onto the issue because one laboratory in his clinical trial consistently reported 20% to 30% lower PSA levels than an outside lab. "That has big implications," potentially the difference between triggering biopsy or giving men a pass until the next year, Vogelzang told MedPage Today.

To see what effect this might be having nationally on diagnosis, the researchers analyzed the population-based National Health and Nutritional Examination Survey (NHANES), in which 5,619 men age 40 and older had PSA testing between 2001 and 2008 using a Hybritech-standardized assay. Gupta and colleagues then converted PSA results for each man to what would have been found with the WHO standardized assays. The median PSA was 0.89 ng/mL by Hybritech assay and 0.73 by WHO assay. They then calculated the number of cancers expected to be diagnosed for men with a Hybritech PSA level above 4.0 ng/mL based on results from the Prostate Cancer Prevention Trial risk calculator, while taking into account the prevalence of PSA screening and compliance rates with biopsy recommendations.

Extrapolated to the national population, 3.53 million undiagnosed men age 40 to 85 would have a PSA at or above the Hybritech threshold of 4.0 ng/mL. But only 59% (2.09 million) of these men would also have a WHO standardized assay-determined level above 4.0 ng/mL. So if clinicians were trying to use the traditional 4.0 ng/mL cutoff with the WHO assay results, they would diagnose prostate cancer in 103,850 per year instead of 157,320 -- missing 34%.

Once physicians are aware of the issue and determine which standard their results follow, they can calibrate their threshold accordingly, Gupta noted. "Knowledge is the only thing that will help," he said.

Action Points
Note that as the result of a largely unrecognized change in the assay by many laboratories to reflect WHO standards, PSA results may be lower than in years past. Researchers in this study recommend adjusting PSA thresholds for biopsy and treatment to reflect the assay standard used by the reporting laboratory.
Note that this study was published as an abstract and presented at a conference. These data and conclusions should be considered preliminary until published in a peer-reviewed journal.

Primary source: ASCO Genitourinary Cancers Symposium Source reference: Gupta A, et al "Impact of the WHO standardization of PSA assays on prostate cancer diagnosis in the United States" ASCO GU 2010; Abstract 100.