Trialling cancer screening alternatives could save lives, not simply find more cancer

THE LATEST REPORT by Public Health Scotland on cancer statistics is available and provides interesting evidence of trends good and bad. I honestly can’t see why people fret endlessly about cancer when they have not made serious attempts to reduce their own risk. It is a bit like fearing a car accident could interrupt your daily session of motorway hopscotch.

Cancer screening is about picking up a lot of cancers cheaply and effectively while, crucially, not picking up things that are not cancer. You want true positives, not false positives. False positives are expensive, consume resources we might better spend elsewhere, and subject people to invasive tests that can harm them physically and psychologically.

The whole point is that when we find cancer, we want a real increase in survival or a meaningful improvement in outcomes. If I find a cancer two years earlier than expected, have I really helped that person live two years longer? Their measured survival from diagnosis may be two years longer, but not necessarily their actual lifespan. They may simply have known for two more years that they had cancer. That is lead-time bias, and it matters enormously when we assess screening.

The prostate is where this becomes particularly difficult. A PSA (prostate-specific antigen) blood test measures the level of a protein made by the prostate gland. Doctors use it to help screen for and monitor prostate cancer, as well as to check for non-cancerous prostate problems. But PSA levels can be raised for reasons other than cancer, may not be dramatically elevated in every clinically important cancer and tend to rise with age. It is not the wonderfully specific cancer detector we might like it to be. An elevated PSA can launch a man into repeat tests, MRI and biopsy; biopsies can miss cancer, and we can end up diagnosing and treating cancers that might never have harmed the patient.

That does not mean prostate cancer is unimportant. Quite the opposite. It is the most common cancer suffered by men in Scotland.

we should want a method that is much better at distinguishing cancers that matter from those that do not

The Scotsman reported, “New statistics from Public Health Scotland show there were 37,657 new cancers registered in Scotland in 2024. For the second consecutive year, prostate cancer was the most common, with diagnoses increasing 6 per cent in a year and 50 per cent in a decade.”

This means that if we are going to screen millions of men, we should want a method that is much better at distinguishing cancers that matter from those that do not.

The UK has now made a difficult but defensible decision. As of the UK National Screening Committee’s March 2026 review, it does not recommend population screening for prostate cancer. It does, however, recommend targeted PSA screening every two years for men aged 45 to 61 with a pathogenic BRCA2 variant and a relevant family history. That is a much more nuanced position than simply saying “we don’t screen”. The remaining question is whether we can find something better for everyone else.

There are already options worth investigating. Why not be daring and run a properly funded UK trial? Scotland, Northern Ireland and Wales have cancer outcomes that differ from England, with deprivation, lifestyle and late presentation among the factors. Give three comparable populations three different approaches: 4Kscore, Stockholm3 and a newer genomic risk model. Then add a fourth arm examining what happens when earlier detection is paired with a better pathway for treating clinically significant early prostate cancer.

Stockholm3 is particularly interesting because it combines protein biomarkers, genetic markers and clinical information. NICE has reported evidence suggesting it can improve detection of clinically significant disease and reduce some unnecessary biopsies compared with PSA-based pathways, while also noting that the evidence is still insufficient to establish its long-term clinical impact. That is precisely why a proper trial is needed.

And measure the right things. Not simply how many cancers we find. Measure clinically significant cancers, unnecessary MRI scans and biopsies, complications, treatment burden, quality of life and, ultimately, whether men live longer or better lives. If one strategy is spectacularly good, adopt it. If they are all rubbish, stop doing them. That is what evidence-based medicine is supposed to look like.

Scotland has a particular reason to try. Public Health Scotland’s latest data for 2020–2024 shows cancer mortality of 397.2 per 100,000 in the most deprived fifth of areas compared with 227.7 in the least deprived. PHS points to differences including cancer type, comorbidity and screening uptake.

If deprivation contributes to poor outcomes, Scotland should become a laboratory for improving them rather than simply a place where we measure the problem yet again.