After a High PSA, a Second Blood Test as Gatekeeper: Reading GOTEBORG-2 and the 4Kscore

An elevated PSA opens a question, not a diagnosis. In a prespecified analysis of the GOTEBORG-2 screening trial, inserting a 4Kscore blood test between an elevated PSA and the MRI and biopsy that usually follow would have spared many men both procedures and cut the diagnosis of low-grade cancers that rarely need treatment, at the cost of delaying an intermediate-grade diagnosis in a small number of men. When you are counseling a worried patient, that single sentence captures the whole bargain: fewer procedures and less overdiagnosis on one side, a modest delay for a few men on the other.

Key points#

• PSA is a good reason to start looking and a poor place to stop deciding.
• A reflex biomarker is a second blood test run automatically after a high PSA, before any imaging, to sort out who really needs to proceed.
• In GOTEBORG-2, a 4Kscore cutoff of 7.5 percent would have avoided many MRIs and biopsies and reduced low-grade (Grade Group 1) overdiagnosis.
• The trade was a delayed intermediate-grade diagnosis in roughly 4 men per 1,000 with an elevated PSA, all of whom stayed in the screening program.
• The figures come from within-trial modeling, not a separate randomized arm, and were drawn from Swedish men aged 50 to 60.

Why a single PSA is a starting line, not a finish line#

PSA measures a protein made by the prostate, and its level rises with cancer but also with benign enlargement, inflammation, and ordinary variation. That makes it a sensitive trigger and a blunt decision tool. A value at or above the widely used 3.0 ng/ml threshold sends a man into a workup that, in current practice, runs from PSA to MRI to a targeted biopsy of anything the scan flags.

The problem is where that pathway leads. A large share of the imaging and biopsies land on men who turn out to have no cancer, or a low-grade cancer that would never have troubled them in their lifetime. Each of those steps carries its own freight: the wait, the worry, the small risks of a biopsy, and the heavy word "cancer" attached to disease that behaves nothing like the cancer your patient is picturing. The clinical question is not whether PSA finds cancer. It is how to avoid acting on the many signals that do not matter while still catching the ones that do.

The idea of a reflex biomarker#

Think of a reflex biomarker as a second gate placed just past the first. When PSA is high, a more specific blood test runs automatically, before the man is ever sent for a scan, and its result decides whether the pathway continues. Men whose biology still looks concerning move forward. Men whose risk of aggressive disease is genuinely low are held back and simply followed.

The 4Kscore is one such gate. It combines four kallikrein proteins in the blood (total PSA, free PSA, intact PSA, and human kallikrein 2) with clinical details into a single number: the estimated percentage risk that a man harbors aggressive prostate cancer. GOTEBORG-2 gave researchers a way to ask what would have happened if that number, rather than the PSA alone, had governed who went on to MRI.

What the trial actually did#

GOTEBORG-2 is a population-based randomized screening trial in Sweden that invited roughly 38,000 men aged 50 to 60. The 4Kscore work, reported by Josefsson and colleagues in European Urology in 2024, was a prespecified, blinded sub-study rather than an afterthought. It drew on men whose PSA was at or above 3.0 ng/ml and who had an evaluable MRI and, where warranted, targeted biopsy results. Blood banked at the time of screening was used to compute a 4Kscore, and the investigators modeled what would have shifted if a cutoff of 7.5 percent had decided who proceeded to imaging.

One design detail governs how much weight you can put on the results. This is a within-trial comparison: the same real, screened cohort measured against its own actual pathway, not a separate arm in which men were prospectively managed by their 4Kscore. That keeps the estimates internally consistent and grounded in a genuine screening population, while leaving the outcomes patients care about most, such as metastasis or death, outside what this particular analysis could measure.

The numbers, per 1,000 men with an elevated PSA#

Framed for every 1,000 men who had crossed the PSA threshold, applying the 4Kscore reflex at 7.5 percent would have:

• avoided MRI for about 408 men (roughly 41 percent of the group),
• avoided biopsy for a further 95 men (about a 28 percent reduction), and
• led to 23 fewer low-grade cancers diagnosed (about a 23 percent drop in Grade Group 1 detection).

Those low-grade cancers are the ones most tied to overdiagnosis. Most are watched with active surveillance rather than treated, and many would never have produced symptoms at all. Reducing how often they are labeled is, in this context, a benefit rather than a miss.

The test also separated aggressive from indolent disease reasonably well. For intermediate-grade and high-grade cancer (Grade Group 2 or higher), the score reached an area under the curve of about 0.84, with a high negative predictive value at the chosen cutoff. Put plainly, a low 4Kscore was a fairly trustworthy signal that clinically significant cancer was unlikely at that moment.

The cost side of the ledger#

The same gate that spares procedures also holds a few men back who would have benefited from moving forward. In this analysis, the reflex step would have delayed an intermediate-grade diagnosis in about 4 men per 1,000 with an elevated PSA, close to 4 percent of that group. Both intermediate-grade cancers that fell below the cutoff in the sample were organ-confined Grade Group 2 disease, found at PSA values just over the threshold.

Delay is not the same as a cancer never found. These men remain in an organized screening program and are retested on schedule, so the question is whether a diagnosis arrives now or somewhat later, not whether it arrives at all. Still, it is a real trade, and the trial states it directly rather than burying it, which is exactly what you should want from evidence like this.

How a generalist should hold the finding#

A few boundaries keep the result in proportion. The men studied were aged 50 to 60 in a single high-participation Swedish program, so the precise percentages should not be lifted wholesale onto older men, different populations, or one-off testing done outside an organized program. The 7.5 percent cutoff is a chosen operating point, and if you nudge it up or down, the balance between procedures avoided and diagnoses delayed slides with it. And because the analysis models a diagnostic pathway rather than tracking survival, it speaks to imaging, biopsies, and grade at detection, not to lives saved decades later.

What GOTEBORG-2 does establish is a direction and a rough size. A reflex kallikrein panel after an elevated PSA can meaningfully reduce MRI and biopsy and can lower low-grade overdiagnosis, while shifting a small number of intermediate-grade diagnoses slightly later. Whether that trade is worth making is a value judgment, weighed by both the health system and the individual man, about how much avoided procedures and avoided overdiagnosis count against the discomfort of a short delay. The practical takeaway is steadier than any single figure, and it is the one to give your patient: one PSA number is a reason to look more carefully, not a verdict to act on at once.