BACKGROUND - Various interferences can cause spurious results for common laboratory tests. Although rare, heterophilic antibodies may produce false elevations in PSA that could prompt unnecessary therapy in men previously treated for prostate cancer. The aim of this study was to determine the prevalence of small, spurious PSA elevations, and the role of heterophilic antibodies.
METHODS - Phase I: all PSA tests drawn and measured between 27 October 2008 and 26 October 2010 at Vanderbilt University Medical Center were analyzed (n=17 133). Patients who had been treated for prostate cancer with PSA values that changed from undetectable to detectable were evaluated. Phase II: patients with a detectable PSA ≤0.5 ng ml(-1) measured between 24 October 2010 and 19 January 2011 were studied prospectively (n=1288). If any patient had a previously undetectable PSA value, their serum was tested for heterophilic antibody interference.
RESULTS - Phase I: 11 men had a spuriously elevated PSA after curative treatment for prostate cancer (0.3%). Mean time to PSA elevation was 3.4±5.5 years, and mean elevation in PSA was 0.33±0.28 ng ml(-1). Each patient's PSA was undetectable after being repeated, and no patient went on to unnecessary treatment. Phase II: 10 men had a newly detectable PSA, 9 of whom had a history of prostate cancer. Each tested negative for interfering heterophilic antibodies when their PSA test was repeated with a heterophilic antibody-blocking reagent.
CONCLUSIONS - In a large cohort, we estimate the prevalence of spuriously elevated PSA values in our population to be 0.3%. No patient with a prostate cancer history was subjected to unnecessary diagnostic evaluation or treatment. On prospective evaluation of PSA conversion to low detectable levels, no patient had evidence of interfering heterophilic antibodies. When using PSA for post-treatment surveillance, it is crucial to confirm all concerning values and consider the presence of a spurious elevation in PSA if the value does not correlate with the clinical scenario.