CADScor's Heart-Failure Data Is Real Science, But the Stock Runs on Its Cash Runway

A press release this week carried a headline most retail investors will read as a growth story opening up: a peer-reviewed study found that the CADScor System — a handheld, acoustic cardiac-diagnostic device made by the Swedish micro-cap Acarix — can diagnose heart failure with far fewer false positives than the standard blood test. Read that way, the little-known company just won a claim to a far bigger market. Read the study the way a doctor or a portfolio manager would, and the true finding is more precise — and the investment question turns out to be about something else entirely.

The machine's edge is not finding the sick, it's not alarming the healthy

Start with what the test actually measures. CADScor is approved in the U.S. and Europe for one job today: rapidly ruling out coronary artery disease at the point of care, in under ten minutes, using acoustics and AI. The new study, published in PLOS Digital Health and led by Aalborg University in Denmark, asked whether that same sensing platform could also flag heart failure — a different condition from CAD, and one of the most common reasons patients end up in a cardiologist's office.

On the core job — identifying who is actually sick — the device did not beat the existing tool. Its diagnostic accuracy, an area-under-the-curve of 92.9%, merely matched the 94.7% of the NT-proBNP blood test that guidelines already recommend. Sensitivity was comparable. So the machine is not "better at heart failure than the blood test"; on the finding part, it is roughly a tie.

Where it clearly won is on the false alarms. The NT-proBNP test, cheap and widely used, is also routinely elevated for reasons that have nothing to do with the heart, so it sends a lot of healthy patients onward. In the study, the blood test flagged a positive predictive value of 27.1% — meaning roughly one in four of its positives were real. CADScor's score more than doubled that, to 57.7%, with specificity of 92% versus the blood test's 68.8%. In practical terms, the signal identified 71 of the 139 referred patients who turned out not to have heart failure, roughly half, as "minimal risk" — patients who, under the current pathway, would likely be booked for an echocardiogram.

And that is the economic mechanism worth pausing on. An echo is expensive, subject to long waits, and operator-dependent. In the U.S. reimbursement story Acarix is trying to build, the pitch to payors is not "our device finds more disease." It is "our device spares you the cost of imaging half the people you'd otherwise image." Fewer false positives is a cost story, and cost stories are what get a small diagnostic device paid.

A research milestone is not a revenue milestone

This is the point where a headline and an operating result diverge, and it matters for anyone deciding what to do with the news. The study is real, peer-reviewed science, but it is deliberately modest in scope. It enrolled 218 patients at two centers in Denmark, and the authors themselves describe the results as exploratory, requiring external validation in larger cohorts before broader clinical use.

More important for the business: none of this is yet sellable. The heart-failure algorithm is investigational. It has been built into the mechanics of Acarix's newer hardware, but it is not part of the system's FDA clearance or its CE mark, which cover only ruling out coronary artery disease. Turning this publication into a product the company can charge for would require a larger validation trial and then an entirely new regulatory clearance. In the language of product cycles, this milestone has not reached any financial result — not one dollar of the revenue this headline might suggest is approved, marketed, or collected.

The variable that sets the near-term outcome is dilution, not the data

Which brings the focus to where it belongs for a stock like this. Acarix is a micro-cap that finances a growing but tiny commercial base through repeated share issuance. In all of 2025 it booked revenue of about SEK 7.4 million — call it under a million U.S. dollars — while losing about SEK 48 million, a loss that narrowed by 27% from the prior year. By the second quarter of 2026 revenue roughly doubled year over year, with the U.S. the leading market, but it remains small. The company ended that quarter with roughly SEK 29 million in cash against a monthly burn of about SEK 3 million — a runway on the order of nine to ten months, absent new capital.

That math is the real clock. A company burning at this rate against this cash balance typically has to raise again before the runway runs out, and Acarix's history is exactly that pattern — a 2024 rights issue saw more than 82 million new shares taken up. Every future raise enlarges the share count, and this is a micro-cap where the number of shares and the size of each raise move the per-share math far more than a promising abstract does.

The reimbursement path it needs to convert the CAD story into recurring revenue is also long and uncertain. Today the U.S. pays a few dozen dollars per test under a temporary code. Acarix has finished enrollment in the trial it needs to move to a permanent code and eventually toward roughly $300 per test, but results are planned only for later this year and meetings with the centers that set Medicare payment rates for early 2027. That is a multi-year, gated sequence that could slip.

So the honest reading of this week's news is a split. As evidence, it strengthens the platform's genuine economic logic — the same case it is already making on CAD, now with data pointing at the still-larger heart-failure cost problem. As a near-term investment driver, it moves almost nothing. What decides whether holding this stock pays is whether Acarix can reach meaningful, reimbursed utilization before the cash runs out and dilution does the work. A validation datapoint is not the same thing as an operating result, and on this balance sheet the difference is the whole story.