New Guidelines for Prostate Cancer Screening

PSA, in conjunction with digital rectal examination, has been used as a screen for prostate cancer since the early 1990s. As a result, fewer men are being diagnosed with advanced disease today and the death rate from prostate cancer has declined significantly. When using a PSA concentration of > or = 4.0 ng/mL in conjunction with DRE as a decision level for considering prostate biopsy, prostate cancer can be diagnosed an average of 5 to 6 years earlier than is possible with DRE alone.¹ This article reviews recent literature that influenced the National Comprehensive Cancer Network (NCCN) to issue new practice guidelines for prostate cancer screening in October 2004. The aim of these guidelines is to help identify at-risk individuals at an earlier age and establish a prostate cancer diagnosis when treatment is more likely to be effective.

NCCN Guideline 1: PSA threshold for considering prostate biopsy lowered to 2.5 ng/mL
The decision to adopt the lower threshold of 2.5 ng/mL was largely based on the results from a recent study by Thompson et al.² Cancer was found in 15.2 percent of 2,950 men with PSA levels < 4.0 ng/mL who underwent prostate biopsy; 14.9 percent of these cases were considered clinically significant. In addition, the investigators found prostate cancer in 26.9 percent of men having PSA values between 3.1 and 4.0 ng/mL. These results are in keeping with a known positive biopsy rate of 22 to 25 percent in men screened for prostate cancer with PSA levels between 2.5 and 4.0 ng/mL,^3,4 and with previous observations that many cancers identified in this lower PSA range are clinically significant.⁵

NCCN Guideline 2: Baseline PSA testing in all men beginning at age 40
PSA levels remain low in younger men without prostatic disease, but gradually increase with age as physiological barriers that keep PSA in the prostatic ductal system become more permeable. Serum PSA levels are even higher in men with benign prostatic hyperplasia (BPH), although the median PSA values from reference range studies are still less than 2.0 ng/mL for men in older age groups.⁶ As Table 1 shows, the average prostate cancer-free man in his fourth or fifth decade of life should have a PSA level that is less than 1.0 ng/mL.

Several investigators have examined the relationship between PSA levels at an early age and the subsequent development of prostate cancer. The richest source of clinical data for this research endeavor comes from the Baltimore Longitudinal Study of Aging (BLSA). This is an ongoing, long-term prospective study conducted by the Gerontology Research Center (National Institute on Aging, Baltimore, MD).⁷ Since its inception in 1958, a total of 1,722 men have participated in this study for varying lengths of time, returning for follow-up visits at approximately 2-year intervals. Serum PSA data are available on more than 1,100 participants, some of whom developed prostate cancer during their follow-up.

Using the BLSA database, Gann et al. evaluated the significance of a single PSA measurement in men (with and without prostate cancer) obtained 10 years before a diagnosis, if any, was made. Results of this study showed that when compared with men with PSA levels below 1.0 ng/mL, men with PSA levels between 2.0 and 3.0 ng/mL were 5 to 6 times as likely to be diagnosed with prostate cancer within 10 years.⁸ In a follow-up study in younger men, prostate cancer risk was established using a PSA value obtained up to 25 years before a diagnosis was established. The relative risk of developing prostate cancer was shown to be more than 3-fold higher in men with PSA levels above the median value for their age group.⁹

The NCCN used age-adjusted reference range values for PSA to define a median value of 0.6 ng/mL for men in their 40s. Since a baseline measurement above 0.6 ng/mL may indicate a higher risk for developing prostate cancer, the NCCN recommends that these individuals undergo PSA testing every 1 or 2 years. Since men with a baseline PSA below 0.6 ng/mL are at a lower risk for prostate cancer, the NCCN recommends no further PSA testing for 5 years (Figure 1). Using such a testing algorithm could theoretically reduce the annual number of PSA tests by 50 percent, saving the healthcare system as much as $1 billion annually, according to Dr. E. David Crawford of the University of Colorado.

Figure 1. NCCN-recommended algorithm for PSA screening.

Dr. Crawford and his colleagues confirmed the safety of this approach in a report presented at the 2002 annual meeting of the American Society of Clinical Oncology. This retrospective study involved 27,863 men (55 to 74 years) who were screened with DRE and PSA measurements. Of the men whose initial PSA value was 1.0 ng/mL or less, 98.7 percent had a value 5 years later that was still < 4.0 ng/mL. Similarly, 98.8 percent of those whose initial reading was between 1.0 and 2.0 ng/mL, still had a value below 4.0 ng/mL two years later.

An added benefit of early prostate cancer detection is the initiation of prompt treatment. Improved outcomes with expeditious surgery¹⁰ and radiation therapy¹¹ have been amply demonstrated, showing that patients with PSA levels below 4.0 ng/mL at the time of detection fare better than most with PSA levels between 4.1 and 10.0 ng/mL.

NCCN Guideline 3: Consider prostate biopsy if the PSA rises at a rate > 0.75 ng/mL/year
The use of PSA velocity (PSAV) as a screening tool for prostate cancer has been well documented in the medical literature. Using the BLSA database, Carter et al. conducted a relatively small study in 1992 describing PSA changes over time in 3 groups of men—normal controls, men diagnosed with BPH and men diagnosed with prostate cancer. Levels of PSA (and serum androgens) were examined during the 7 to 25 years prior to histologic diagnosis or exclusion of prostate disease. Results of this study showed that age-related changes in serum androgen levels were not significantly different between groups; neither were the PSA levels in the men with BPH and prostate cancer measured 5 years before diagnosis. However, using a PSAV cutpoint of 0.75 ng/mL per year, the authors showed a high specificity (95%) for distinguishing cancer from nonmalignant cases when repeated PSA measurements over an appropriate time span were used for PSAV calculations.¹²

Differences in PSAV between men with and without prostate cancer, whose initial (prediagnosis) PSA value was between 2.0 and 4.0 ng/mL, have subsequently been confirmed during a 10-year follow-up period.¹³ This study, also using the BLSA database, showed the relative risk of developing prostate cancer to be 6.5-fold higher in men having a PSAV > 0.1 ng/mL/year, and reported a 44 percent incidence of prostate cancer when the PSAV exceeded 0.2 ng/mL/year.

The NCCN recommends that PSA values used to calculate PSAV should be performed by similar assay techniques in the same clinical laboratory and that velocity should be calculated from at least three PSA values obtained over at least an 18-month period. This is necessary to minimize the potential influence of physiological and analytical variability upon measured values.

NCCN Guideline 4: Extended-pattern biopsy technique improves the rate of positive biopsy
The NCCN points out that approximately 75 to 80 percent of patients with a PSA level > 2.5 ng/mL will not be found to have cancer on first biopsy. However, at least 10 percent of those patients with an initially negative biopsy may prove to have prostate cancer on subsequent biopsy. Table 2 describes the NCCN-recommended course of action for different PSA and/or free PSA levels following an initial negative prostate biopsy result.

To decrease the likelihood of a false negative biopsy, the NCCN encourages a more extensive sampling technique, both initially and for repeat biopsies. In addition to the standard six ("sextant") cores, NCCN recommends an additional four cores be directed at the lateral peripheral zone and to any palpable or suspicious area(s) visualized on ultrasonography. Transition-zone biopsy is not supported for a routine biopsy, but may be considered in a repeat biopsy if PSA is persistently elevated. The NCCN notes that there is a decreased probability of finding cancer after a second negative extended-pattern biopsy. For high-risk men with multiple negative biopsies, consideration can be given to a saturation biopsy strategy.

Summary and conclusions
The literature supporting the new NCCN guidelines indicates that the long-term risk of developing prostate cancer in at-risk men within a two- to three-decade span is a function of the PSA level, and that the PSAV in cancer cases will be greater than that of patients with BPH or no evidence of prostate disease. The NCCN points out that patients at higher risk for prostate cancer (e.g., those having a first-degree relative with prostate cancer or being of African-American descent) may benefit from earlier baseline testing.

Not all prostate cancers diagnosed in younger men will require immediate treatment; therefore a period of "watchful waiting" is normally indicated. During this interval, the biology of the cancer can be determined by serial PSA testing (PSAV), and tissue or serum samples can be checked for the presence of other markers having prognostic significance. It is estimated that up to 30 percent of cases of PSA-detected prostate cancer are indolent tumors that pose no threat,¹⁴ at least not at the time of diagnosis. Ultimately, the choice of initiating or forgoing treatment requires a careful assessment of all of the clinical and laboratory factors that may influence outcome.

The excellent precision of DPC's Third Generation PSA assay at total PSA values below 1.0 ng/mL makes it well suited for use within the context of the new NCCN guidelines. Not only can this assay serve as the laboratory's routine PSA test, but its superior low-end sensitivity provides earlier detection of recurrent prostate cancer than is possible with less sensitive PSA assays.