Felix Ortiz: Tess, hey — did you sleep this week or did you just mainline medical journals?
Tess Hollis: Bit of both, honestly. I kept hitting this one number and not being able to put it down.
Felix Ortiz: Yeah, same — I have the same problem. We're getting into prostate cancer screening, and one in eight men will be diagnosed in their lifetime. That's the stat everyone leads with. And I kept thinking, yeah but — wait, what does that actually mean for any one person sitting in a waiting room right now?
Tess Hollis: That's the question. And the American Cancer Society just put a number on it — 333,830 new cases projected for 2026 in the U.S. alone. That's not a lifetime stat. That's this year.
Felix Ortiz: This year. Okay — so picture this. It's a Tuesday afternoon. Guy named David, fifty-eight years old, he's sitting in a urology waiting room. Fluorescent lights, those little brochure racks with laminated pamphlets about PSA levels — Prostate-Specific Antigen, the blood test that's been the standard since the FDA approved it in 1994. He got a referral from his GP because his PSA came back elevated. Not crazy high, just — enough. And he's holding this referral slip and I don't know, I keep thinking about what's going through his head.
Tess Hollis: Right — and then the urologist walks in.
Felix Ortiz: And the urologist walks in and says they want to do a multiparametric MRI — an mpMRI. And that's the moment, right? That's where David's Tuesday becomes something else entirely.
Tess Hollis: And that's complicated.
Felix Ortiz: Complicated how, though? Like, statistically complicated or — what are we actually saying about David's odds here?
Tess Hollis: Think of it like a net. You drag it through the ocean to catch one specific fish. And you do catch it — but you haul up 499 others in the process, and sorting through those 499 has a real cost.
Felix Ortiz: Wait — 499 isn't a metaphor number, is it.
Tess Hollis: No, that's — actually that's almost exactly the Cochrane number. They pulled together six trials, nearly 800,000 participants. And the 23-year follow-up alone covered 162,241 men. Cochrane's conclusion, moderate certainty: PSA-based screening prevents roughly 2 deaths per 1,000 men screened. Which means one death prevented per 500 men who go through the whole process.
Felix Ortiz: So David — statistically — is almost certainly in the 499.
Tess Hollis: Right. And the tradeoff sitting inside that number — overdiagnosis, overtreatment — that's not a side effect, that's the central problem. Indolent cancers that would never have caused symptoms, never caused death, getting caught, getting treated anyway.
Felix Ortiz: And the treatment isn't nothing — like, we're not talking about a pill.
Tess Hollis: Exactly, it's — yeah, surgery, radiation, the side effects that come with those. So the question isn't just whether screening finds cancer. It's whether we can tell David he's in the 499 before the biopsy, not after. That's what's actually driving the whole push toward mpMRI and the new biomarkers — can we sort the net faster, before the cost lands on the patient.
Felix Ortiz: So the new tools aren't about finding more cancer. They're about finding less of the wrong cancer.
Tess Hollis: That's exactly it — and that's where the AUA and SUO just stepped in. February 26, 2026, they dropped an amendment to the Early Detection of Prostate Cancer Guideline. New statements on imaging, biomarkers, biopsy technique. It's a whole updated playbook.
Felix Ortiz: Yeah — and the toolkit they're pointing to is genuinely different. Like, mpMRI, biparametric MRI which is the faster version that skips the contrast injection, ClarityDX Prostate which is a multivariable blood test — it was studied in npj Digital Medicine for predicting clinically significant cancer. PCA3 urine test. AI-assisted reading under PI-RADS. It's a lot. But I'm stuck on whether any of it actually closes the accuracy gap.
Tess Hollis: And even the MRI — interreader agreement is only moderate?
Felix Ortiz: Still moderate. PI-RADS standardized the reporting language, but two radiologists looking at the same scan still disagree more than you'd want.
Tess Hollis: So we layered AI on top of a process that was already inconsistent. AI targeting the false negatives, the false positives, the long scan times — but the foundation is still shaky.
Felix Ortiz: Right — and PCA3, the urine test? Moderately accurate is basically the descriptor. That's the field's own language.
Tess Hollis: And then there's DRE. Digital rectal exam. Still in some guidelines — and the pooled sensitivity across seven community screening studies is 0.51.
Felix Ortiz: That's a coin flip.
Tess Hollis: Literally. And the guideline amendment updates all this — imaging, biomarkers, technique — but a guideline amendment isn't a practice change. A radiologist in a rural hospital isn't running mpMRI the week after a PDF drops in D.C.
Felix Ortiz: Which is — yeah, actually that connects to something we haven't touched yet. Because what counts as the cancer you actually need to find versus the one you'd be better off never knowing about — that question gets a lot messier when you hear what happened to one specific patient over five years of biopsies. That one's coming.
Tess Hollis: That patient story is exactly where the 1-in-8 number breaks down. The American Cancer Society puts a man's lifetime prostate cancer diagnosis risk at 1-in-8 — and that makes it the most common solid organ cancer in men. But the number isn't telling you what kind of cancer.
Felix Ortiz: What do you mean, what kind?
Tess Hollis: Clinically significant prostate cancer — csPCa — that's the aggressive kind. The one that causes harm or death if you don't treat it. That's the actual target. But a huge fraction of what gets caught is indolent disease — low-risk, may never cause symptoms, may never surface in a man's lifetime.
Felix Ortiz: Wait, so 1-in-8 is a diagnosis risk — not a 'this will hurt you' risk.
Tess Hollis: Right. And the STAT News piece from January 21st, 2026 — a patient wrote it himself — he gets diagnosed, and then the cancer just... disappears. Not in a medical-miracle sense. It's gone on the next biopsy. And the one after that. Five subsequent annual biopsies, no cancer reappears.
Felix Ortiz: But he was diagnosed. He had to live with that. For years.
Tess Hollis: That's the harm. Not the biopsy needle — the diagnosis. And the 98% ten-year survival rate that sounds so reassuring? We can't actually separate that from lead-time bias. We might just be catching cancers earlier and calling the clock a cure. That's why shared decision-making is the framework the AUA and SUO amendment leans on explicitly — because 2 deaths prevented per 1,000 is a population number. It does not tell that man what his biopsy means for him specifically.
Felix Ortiz: Yeah and — wait, that's the thing nobody says in the waiting room, is it. The pamphlet doesn't say 'your cancer might not be the dangerous kind.'
Tess Hollis: Which is exactly what David's sitting with right now. The 2026 AUA and SUO amendment — it doesn't give the doctor a single universal threshold anymore. It pushes the clinician to actually say out loud: here are your individual risk factors, here's what the mpMRI showed, here's what ClarityDX is telling us — and then ask the patient what matters to him and what he's willing to risk to find out.
Felix Ortiz: That's — yeah, that's a harder conversation than 'your PSA is elevated, let's biopsy.' Like, the doctor now has to actually ask what the patient values. That's not nothing.
Tess Hollis: And the real test of whether any of this works — mpMRI, ClarityDX, AI-assisted imaging — it's not whether it performs in a well-resourced academic center. It's whether a man who has historically been screened less and died more from prostate cancer can actually get there. Because if the precision tools stay concentrated where access already exists, the framework just widens a gap that's already bad.
Felix Ortiz: David's still in that chair. That's what really matters to me. Not a verdict. Not a pamphlet stat. Just a guy waiting for a doctor to ask him the right question.
Tess Hollis: Yeah. That part hasn't changed.
