Overview of #388 — Prostate cancer screening (Peter Attia, MD)
This episode is a deep dive into prostate cancer screening—why population-level PSA guidelines have failed, how modern tools and workflows substantially improve early detection, and practical implications for patients and clinicians in 2026. Dr. Peter Attia explains the biology of PSA, the historical harms that drove guideline changes, the data flaws that underpinned those changes, and a contemporary, evidence-based screening algorithm (PSA velocity + MRI + transperineal biopsy + active surveillance). He closes with an urgent warning about PSA suppression from finasteride and clear, actionable recommendations.
Key takeaways
- PSA screening done poorly (single-value testing → reflex transrectal biopsy → overtreatment/infection) contributed to harms and to the 2012 USPSTF recommendation against routine screening.
- With modern practice (longitudinal PSA/velocity, PSA density from imaging, biparametric MRI, transperineal biopsy, and active surveillance) the balance of benefits vs harms strongly favors screening.
- The PLCO trial—the main negative trial used to justify stopping screening—was heavily “contaminated” (control arm had widespread PSA testing); corrected analyses of the same data show a 27–32% mortality reduction with screening.
- Finasteride (and dutasteride) dramatically suppress PSA and can mask cancer unless clinicians correct for its effect; failure to do so has been associated with higher rates of metastatic diagnosis.
- Practical, low-cost steps can catch many lethal prostate cancers early and could meaningfully reduce prostate cancer mortality.
PSA: biology and what it actually measures
- PSA = prostate-specific antigen, a protein produced by prostate tissue; small amounts normally leak into blood.
- PSA correlates with prostate volume (benign growth increases PSA), and with disruption of gland architecture (malignancy increases PSA leakage).
- Single PSA values are noisy (day-to-day variability ≈ ±15%; ejaculation can spike PSA by ≈40%). A single snapshot is limited—trend over time (PSA velocity) is far more informative.
Why past screening guidance went wrong
- 1990s–2000s workflow: elevated PSA → ultrasound → transrectal biopsy → treat any cancer found. Problems:
- High false positives (benign prostatic hyperplasia, prostatitis, sexual activity, natural variation).
- Transrectal biopsies carried infection risk (5–7%) and often required antibiotics.
- Routine immediate treatment of low-risk cancers led to overtreatment (ED, incontinence, radiation toxicity, psychological harm).
- USPSTF used PLCO and similar data to argue harms > benefits; however PLCO control arm was heavily contaminated—most "no-screen" men actually received PSA testing.
- Reanalysis correcting for contamination: PLCO supports a ~27–32% reduction in prostate cancer mortality with screening.
Modern screening toolbox (how practice has changed)
1) PSA as a longitudinal measure (PSA velocity)
- Focus on trend across serial tests rather than single value.
- Example thresholds Dr. Attia cites:
- For men with baseline PSA <4 ng/mL: sustained rise >0.35 ng/mL per year over ~18 months is a red flag.
- For men with baseline PSA ≥4 ng/mL: threshold ≈0.75 ng/mL per year.
- Practical example: 55-year-old with PSA 1.3 → 1.9 over 18 months = velocity ~0.4/year → trigger further workup.
2) MRI and PSA density
- MRI provides prostate volume (for PSA density) and identifies suspicious lesions noninvasively.
- PSA density = PSA / prostate volume; disproportionate PSA relative to volume suggests malignancy.
- PRIME trial (2025): contrast-free biparametric MRI performed essentially identically to full multiparametric MRI for detecting clinically significant cancers (143 vs 145 detections out of 490). Implication: faster, cheaper, more scalable pre-biopsy MRI without contrast is effective.
3) Safer, more accurate biopsy: transperineal approach
- Transrectal biopsy: infection risks (5–7%), sampling bias toward posterior prostate.
- Transperineal biopsy: entry through perineum, avoids rectum → dramatically lower infection risk (0% infection in a multicenter phase 3 trial of ~1,700 patients) and better access to anterior/apical prostate; detects more high-grade cancers and fewer low-grade ones.
- Adoption lagging: ~37% of U.S. urologists currently perform transperineal biopsies.
4) Active surveillance and nuanced treatment decisions
- Pathology uses Gleason scoring → grade groups (1–5). Low-grade (Gleason 3+3 or grade group 1–2) often indolent.
- Modern practice: many low-grade cancers move to active surveillance (serial PSA, MRI, selective biopsy), reserving treatment for progressing or higher-grade tumors.
- This reduces overtreatment and shifts screening benefit/harm calculus toward screening.
The PLCO data problem (short summary)
- PLCO control arm had substantial PSA screening contamination (40–60% per year; >90% had at least one PSA by the end).
- Effect: a study comparing screening to screening concluded no benefit.
- Corrected reanalysis: when accounting for contamination/lead time, PLCO shows ~27–32% prostate cancer mortality reduction from routine PSA screening.
- Dr. Attia argues major guideline groups (USPSTF, Canadian Task Force, AAFP) should revisit recommendations given modern evidence and reanalyses.
Finasteride/dutasteride: an urgent clinical pitfall
- 5-alpha reductase inhibitors (most commonly finasteride) reduce intraprostatic DHT and suppress PSA production.
- Typical PSA effects:
- ~33% reduction in PSA within 6 months; ~50% by 1 year.
- Suppression persists while on the drug.
- Clinical correction factors:
- FDA recommends doubling PSA in men on finasteride.
- Dr. Attia’s practice: multiply PSA by 2.3 (2–7 years on drug) and 2.5 (>7 years).
- PSA velocity on finasteride: after ~12 months, PSA should not rise; any upward trend is highly suspicious.
- Real-world consequence: VA study showed metastatic diagnosis rates higher in men on 5-ARIs (6.7% vs 2.9%)—likely because raw PSA values were read without correction.
- Action: clinicians must ask about 5-ARI use and adjust PSA interpretation; patients must disclose finasteride use.
Concrete recommendations (for patients and clinicians)
- For clinicians:
- Move from one-off PSA testing to longitudinal monitoring (calculate PSA velocity).
- If velocity is concerning, order pre-biopsy biparametric MRI (PSA density + lesion identification).
- If biopsy is indicated, prefer transperineal, MRI-targeted approach when possible.
- Use active surveillance for low-grade disease; treat higher-grade disease promptly.
- Routinely ask about finasteride/dutasteride and apply appropriate PSA correction factors.
- Re-evaluate guideline-based practices in light of PLCO reanalysis and modern diagnostic tools.
- For patients:
- Ask your PCP about baseline PSA and serial testing—PSA is inexpensive and often included in routine bloodwork.
- If on finasteride/dutasteride, tell your clinicians and insist PSA be adjusted for the drug’s effect.
- If PSA velocity, PSA density, or MRI is concerning, request MRI and, if needed, a transperineal targeted biopsy.
- If diagnosed with low-grade prostate cancer, discuss active surveillance as the first-line management.
Important statistics & illustrative numbers
- Early (stage I–III) 5-year survival: >99%; 15-year survival from early detection ~97%.
- Stage IV (metastatic) 5-year survival: ≈38%; median survival ≈30 months.
- Annual US prostate cancer deaths: ≈36,000 (current context from episode).
- PSA variability: day-to-day ±15%; ejaculation can increase PSA ~40%.
- PRIME trial: biparametric MRI vs full multiparametric MRI detection difference ≈0.4 percentage points (143 vs 145/490).
- Biopsy infection: transrectal 5–7% vs transperineal 0% in large trial (>1,700 patients).
- PLCO reanalysis: 27–32% mortality reduction with proper accounting for contamination.
- Finasteride correction: typical PSA falls ≈33% in 6 months; multiply PSA roughly ×2 (FDA) — Attia uses ×2.3 (2–7yrs) and ×2.5 (>7yrs).
Final framing (Attia’s conclusion)
- Prostate cancer follows a detectable, often slow path from localized to metastatic disease—making screening inherently valuable.
- Modern diagnostics and management (velocity + PSA density + biparametric MRI + transperineal biopsy + active surveillance) mitigate earlier harms and make screening far more effective and safer than 15 years ago.
- The barrier today is not the lack of tools but awareness, outdated guidelines, and misinterpretation (especially in men on finasteride). If implemented broadly and intelligently, many prostate cancer deaths could be prevented.
For deeper detail, Dr. Attia points listeners to the episode show notes (studies and tables) and a companion newsletter that outlines his practice’s PSA approach.