Screening for prostate cancer using PSA with and without MRI: systematic reviews with meta-analysis

This systematic review and meta-analysis of 15 randomized controlled trials and 8 observational studies concludes that while repeated PSA-based screening for men aged 55–69 likely reduces prostate cancer mortality and metastasis at the cost of significant overdiagnosis and false positives, adding MRI to the screening pathway effectively reduces false positives and overdiagnosis without compromising the detection of clinically significant cancers, though long-term mortality benefits of the MRI strategy remain unproven.

The Gist

Imagine your body is a vast, quiet forest. Most of the time, it's peaceful. But sometimes, a small, slow-growing weed (prostate cancer) starts to sprout. Some of these weeds are harmless and will never bother the forest; others are aggressive and could cause real damage.

For years, doctors have used a tool called the PSA test to scan the forest. It's like a metal detector that beeps when it finds something metallic. The problem? The metal detector beeps for harmless rocks, old cans, and actual treasure alike. It creates a lot of "false alarms," leading to unnecessary digging (biopsies) and sometimes even removing harmless plants (overdiagnosis).

This new study is like a massive, 20-year-long expedition where researchers looked at thousands of forests to answer two big questions:

1. Is it worth using the metal detector (PSA) at all?
2. If the detector beeps, should we use a high-tech drone (MRI) to look closer before we start digging?

Here is what they found, broken down simply:

1. The Metal Detector (PSA Screening) vs. Doing Nothing

The researchers looked at men starting screening at different ages. Think of this like checking your forest at different times of the day.

• Starting too early (Age 50-54): The forest is still young. The metal detector finds a few bad weeds, but mostly it just finds rocks. You end up doing a lot of unnecessary digging. The benefit is small; the hassle is big.
• Starting in the "Sweet Spot" (Age 55-69): This is the best time. If you check the forest every 2–4 years during these years, you are likely to find and remove the dangerous, fast-growing weeds before they destroy the forest.
  • The Good News: For every 1,000 men who get screened, about 2 fewer men will die from prostate cancer, and about 6 fewer will get the "invasive" kind of cancer that spreads.
  • The Bad News: To save those few lives, you have to dig up the soil for about 150 men who didn't need it (false alarms), and you will accidentally remove about 24 harmless weeds (overdiagnosis) that would have never caused trouble.
• Starting too late (Age 70+): The forest is old. The weeds have either been there for a long time or won't grow fast enough to matter before the tree falls naturally. Checking now doesn't really help save lives, but it still causes stress and unnecessary digging.

The Takeaway: Screening is a trade-off. You save a few lives, but you cause a lot of anxiety and minor procedures for many others. It's not a clear "yes" or "no"; it's a "maybe, depending on how much risk you're willing to take."

2. The High-Tech Drone (Adding MRI)

Now, imagine the metal detector beeps. In the old days, you would immediately start digging (biopsy). This study suggests a smarter approach: Send in a drone (MRI) first.

• How it works: If the PSA test beeps, instead of digging immediately, you take a high-resolution photo of the spot with an MRI.
• The Result: The drone can tell the difference between a harmless rock and a real weed.
  • Fewer False Alarms: The study found that using the drone first prevents about 33 unnecessary digs for every 1,000 men.
  • Fewer Unnecessary Removals: It stops doctors from removing about 10 harmless weeds (low-risk cancers) that would have just sat there quietly.
  • Same Protection: Crucially, the drone doesn't miss the dangerous weeds. It finds the bad ones just as well as the old method, but without the mess.

The Catch: Drones are expensive, need special pilots (radiologists), and take time. Not every town has one. So, while it's the "perfect" way to do it, it's not always practical everywhere yet.

3. What About Checking Every Year?

Some people thought checking the forest every single year would be better. The study says no. Checking every 2–4 years is just as good at finding the bad weeds, but it saves you from the stress and cost of checking too often.

The Big Picture: A Balancing Act

This study is like a scale.

• On one side, you have Lives Saved and Cancer Prevented.
• On the other side, you have Anxiety, Unnecessary Procedures, and Over-treatment.

The study tells us that for men aged 55 to 69, the scale can tip in favor of screening, but only if you are okay with the fact that many of your friends will get a "false alarm" and a scare.

In simple terms:

• Don't panic if you are under 55 or over 70; the test probably won't help you much.
• If you are 55–69, talk to your doctor. If you choose to screen, do it every few years, not every year.
• If the test is positive, ask for an MRI before agreeing to a biopsy. It's like getting a second opinion from a satellite before you start digging holes in your backyard.

This research helps us stop guessing and start making decisions based on real numbers, so we can protect the forest without tearing it apart.

Technical Summary

1. Problem Statement

Prostate cancer screening using Prostate-Specific Antigen (PSA) has historically been controversial due to the trade-off between reducing mortality and causing significant harms, specifically overdiagnosis (detecting indolent cancers that would never cause symptoms) and false positives (leading to unnecessary biopsies).

• Limitations of Previous Evidence: Prior guidelines (e.g., the 2014 Canadian Task Force on Preventive Health Care) relied on trials with insufficient follow-up duration to fully assess mortality benefits and overdiagnosis rates in the context of slow-growing prostate cancer.
• Knowledge Gaps: There was a lack of data on:
  • Absolute effects stratified by age and screening intensity (frequency/number of rounds).
  • The efficacy of sequential screening (PSA followed by MRI for positive results) in reducing harms without compromising the detection of clinically significant cancer.
  • Evidence certainty assessments using modern GRADE methodologies and decision thresholds based on absolute effect sizes.

2. Methodology

The study followed a rigorous protocol registered with PROSPERO (CRD420250651056) and adhered to PRISMA and AMSTAR 2 standards.

• Data Sources & Search:
  • Searched MEDLINE, Embase, and CENTRAL from 2014 to January 28, 2026.
  • Included pre-2014 studies on PSA/DRE screening from existing systematic reviews.
  • Study Types: Randomized Controlled Trials (RCTs) and prospective observational studies.
• Key Questions (KQs):
  • KQ1: Screening (PSA ± DRE) vs. No Screening.
  • KQ2: Sequential screening (PSA + MRI triage) vs. PSA alone.
  • Population: Men not known to be at high risk (e.g., no germline mutations).
• Study Selection:
  • Included 15 RCTs ($N=856,000$) and 8 observational studies ($N=56,122$).
  • The European Randomized Study of Screening for Prostate Cancer (ERSPC) was treated as 8 separate trials due to site-specific variations in protocols.
  • AI-assisted screening (DistillerAI) was used for title/abstract screening.
• Analysis & Synthesis:
  • Meta-analysis: Random-effects models were used to pool data.
  • Subgroup Analysis: Stratified by age (50–54, 55–69, 70–74), screening intensity (low vs. moderate/high), and risk of bias.
  • Certainty Assessment: Used GRADE (Grading of Recommendations Assessment, Development and Evaluation).
  • Decision Thresholds: Pre-defined absolute effect thresholds were established to determine clinical importance (e.g., $\ge$1 fewer death per 1,000 men). Relative effects were converted to absolute effects using Control Event Rates (CERs) derived from Canadian statistics and trial data.
  • Time Horizon: 20-year follow-up for ages 50–69; 15-year follow-up for ages 70–74.

3. Key Contributions

• Long-term Follow-up: Provides the first comprehensive synthesis of screening data with approximately two decades of follow-up, addressing the "lag time" issue in previous reviews.
• Absolute Effects by Age: Moves beyond relative risk (RR) to provide absolute risk differences per 1,000 men, stratified by starting age, which is critical for shared decision-making.
• MRI Integration: Offers the first systematic evaluation of adding MRI to PSA screening as a triage tool, quantifying its impact on reducing false positives and overdiagnosis.
• Risk of Bias Granularity: Treated ERSPC sites as distinct trials to account for heterogeneity in screening intensity and contamination rates, providing a more nuanced risk-of-bias assessment.

4. Key Results

A. PSA Screening vs. No Screening (KQ1)

• Mortality Benefit:
  • Ages 55–69 (Moderate/High Intensity): Screening every 2–4 years likely reduces prostate-cancer mortality by $\ge$2 fewer deaths per 1,000 men over 20 years (Moderate certainty).
  • Ages 50–54: Benefits are smaller (likely $\ge$1 fewer death per 1,000) with similar harms.
  • Ages 70–74: No important benefit or harm identified (Low to Moderate certainty).
• Metastatic Cancer: Screening starting at 55–69 likely reduces metastatic cancer incidence by $\ge$6 fewer cases per 1,000 (Moderate certainty).
• Harms (Overdiagnosis & False Positives):
  • Overdiagnosis: Screening starting at 55–69 likely causes $\ge$24 additional cases of overdiagnosis per 1,000 men (Moderate certainty). Most overdiagnosed cases are low-risk (Gleason 6).
  • False Positives (FPs): Screening multiple times results in $\ge$150 FPs per 1,000 men.
  • Biopsy Complications: $\ge$230 complications per 1,000 men screened multiple times.
• Screening Intensity & Modality:
  • Annual Screening: Does not add benefit compared to 2–4 year intervals.
  • Adding DRE: Does not add benefit to PSA alone.
  • Low Intensity (1 round): Uncertain benefit; likely does not reduce mortality by the threshold of 1 per 1,000.

B. Sequential Screening (PSA + MRI) vs. PSA Alone (KQ2)

• Clinically Significant Cancer: Adding MRI to triage positive PSA results makes little-to-no difference in the detection of clinically significant cancer (Gleason $\ge$7).
• Clinically Insignificant Cancer: Sequential screening reduces the detection of insignificant cancers by $\ge$10 fewer cases per 1,000 (High certainty).
• False Positives: Sequential screening reduces FPs by $\ge$33 fewer per 1,000 (High certainty).
• Limitations: Findings are limited to one round of screening; long-term mortality data and age-specific stratification for this strategy are currently unavailable.

5. Significance and Interpretation

• Complex Trade-offs: The study confirms that while PSA screening reduces prostate-cancer mortality and metastatic disease, the benefit is accompanied by substantial harms (overdiagnosis and false positives). The decision to screen requires weighing a reduction of ~2 deaths against ~24 overdiagnoses and ~150 false positives per 1,000 men.
• Age-Dependent Recommendations: The benefit-to-harm ratio is most favorable for men starting screening between ages 55 and 69 with moderate-to-high intensity (every 2–4 years). Screening starting at age 50–54 offers smaller benefits, and screening starting at 70+ offers no clear net benefit.
• Role of MRI: The addition of MRI for men with elevated PSA is a promising strategy to optimize the benefit-harm profile by significantly reducing false positives and overdiagnosis without missing significant cancers. However, implementation requires consideration of costs, infrastructure, and equity.
• Guideline Implications: These findings challenge previous guidelines that may have been based on shorter follow-up periods. They support a shift toward age-stratified, shared decision-making rather than blanket population screening. The results suggest that annual screening and the addition of DRE are not cost-effective or beneficial.

Conclusion: This review provides the most robust evidence to date on the absolute effects of prostate cancer screening. It highlights that while screening saves lives, the trade-offs are significant. The integration of MRI as a triage tool offers a pathway to mitigate harms, though long-term data on mortality with this approach is still needed.