How helpful is prostate cancer screening using PSA?

Whether prostate-specific antigen (PSA) screening is genuinely useful is one of the most divisive debates in modern oncology. Detecting cancer early to prevent death is intuitively appealing, but the controversy arises from a tradeoff between mortality reduction and the harms of overdiagnosis. While PSA testing captures aggressive cancers, it also frequently flags slow-growing, indolent tumors that would never have threatened the patient’s life. This lack of specificity has historically led to unnecessary biopsies and overtreatment, subjecting numerous men to life-altering side effects like incontinence and impotence for a disease that might have remained harmless. As a result, the debate has centered not on whether screening works technically, but on whether the price paid in quality of life by the many justifies the survival benefit gained by the few.

Prostate cancer is one of the most common cancers in men worldwide, yet because it often grows slowly, many men die with the disease rather than from it. Despite this, the threat is substantial, with global mortality projected to double by 2040 as life expectancies rise. For decades, the frontline defense has been the prostate-specific antigen (PSA) test, a simple blood screen that measures a protein produced by the prostate gland intended to flag early signs of malignancy. Elevated levels can indicate cancer, but they are also frequently caused by benign conditions such as an enlarged prostate or infection. A positive result typically triggers a diagnostic cascade, moving from the blood draw to more advanced tools like magnetic resonance imaging (MRI) and, ultimately, an invasive needle biopsy.

The latest update from the European Randomized Study of Screening for Prostate Cancer (ERSPC) provides valuable new data that address this thorny question. Starting in the early 1990s, researchers have been tracking men for more than two decades. An interim report at the 16-year mark had already established that screening could save lives -- showing a 20 percent relative reduction in mortality -- but that victory came at a steep cost of overdiagnosis. The new report with 23 years of follow-up data offers the most comprehensive look yet at whether the long-term survival benefits justify the immediate harms.

The study enrolled a group of over 162,000 men aged 55 to 69 across seven European nations. Initiated in 1993, the trial randomly split participants into two camps: one invited for repeated PSA screening, typically at four-year intervals, and a control group that received no such prompts. Men whose test results hit a specific threshold, most often 3.0 nanograms per milliliter, were referred for biopsies to check for malignancy. By linking these participants to national registries for a median of 23 years, the researchers were able to track cancer diagnoses and deaths well beyond the active screening period. It should be noted that there was minor "contamination" between the groups with a small number of control men also receiving the PSA tests (perhaps as part of a checkup), which could dilute the observed benefits (and/or harms) of the screening group.

Researchers calculated the number of cases in both groups and found that the cumulative incidence of prostate cancer was higher in the screening group (14%) compared to the control group (12%). This is as expected because the point of screening is to detect cancers that may go unnoticed because they escape other diagnostic criteria such as overt symptoms including urinary difficulties (e.g. pain, frequency), blood in urine/semen, and sexual dysfunction. The difference was termed excess cases of prostate cancer diagnosed, and there were 27 excess cases in the screening group compared to the control group per 1000 men.

Importantly in the screening group, the number of low-risk cancers was roughly twice as frequent (Rate Ratio: 2.14), whereas the number of advanced cancers was significantly less frequent (Rate Ratio: 0.66). Most likely, the PSA test flagged people with prostate cancer early when the cancer was still low risk (i.e. lower PSA level, biopsy not as abnormal, more localized, etc.) which enabled earlier treatment, and for some patients in the screening group, this intervention prevented the cancer from progressing to an advanced stage. For other patients this intervention in hindsight proved to be unnecessary because the cancer was growing so slowly.

The bottom line is ultimately mortality, and at 23 years, prostate cancer mortality was 1.4% in the screening group versus 1.6% in the control group (Figure 1). This translates to a 13% relative reduction (~0.2/1.6) in the risk of death from prostate cancer (Rate Ratio: 0.87; 95% CI, 0.80 to 0.95). When adjusted for nonattendance (not all patients in the screening group who tested positive went ahead with the biopsy), the risk ratio improved to 0.84.

One can also compare the results at the 16-year mark of the trial compared to the 23-year mark. It appears that the mathematics of the tradeoff have improved over time. The absolute risk reduction continued to rise (0.22% at 23 years vs. 0.14% at 16 years (Figure 1). The harm-to-benefit profile has also become more favorable over time, indicated by a significant decrease in the "number needed to invite" (from 628 down to 456) and "number needed to diagnose" (from 18 down to 12) to prevent one death. Finally, the true benefit may still be underestimated because many of the youngest participants have not yet reached the age (80+) where prostate cancer mortality typically peaks.

The authors conclude that while the lifesaving potential of screening is now undeniable, the strategy must evolve to curb its collateral damage. They emphasize that the "harm-to-benefit" profile has improved over the decades, yet the sheer volume of unnecessary biopsies and the overdiagnosis of harmless tumors remain critical concerns. The way forward, they suggest, lies not in abandoning the PSA test but in refining it with "risk-based" approaches -- using tools like M.R.I. and algorithmic calculators to filter out false alarms before invasive procedures occur. For example, men with very low baseline PSA levels (≤1 ng/mL at age 60) have a very low long-term risk of death from the disease, and perhaps do not need to be tested as frequently.

Finally, the researchers optimistically note that because diagnostic and treatment technologies have advanced significantly since the study began in the 1990s, the outlook for patients today is likely even brighter than these historical numbers suggest.

Figure 1. The mortality benefit of PSA screening. Cumulative mortality over 23-year period in screening group (blue line) and control group (gray line). The screening group exhibits a lower death toll from prostate cancer, and the gap is slowly growing over time. These data show that early detection saves lives despite the risk of overdiagnosis (Roobol et al. NEJM, 2025).