NT-proBNP — how much a blood test can say about the heart

NT-proBNP — how much a blood test can say about the heart

"If only a blood test could check the heart" — every caregiver has the thought. The echo is expensive, needs booking, and the cat panics at the clinic door.

The test that comes closest is NT-proBNP. It is genuinely useful and has already appeared in supporting roles across this bible. But it is a test with sharply divided strengths and weaknesses — it is worth exactly as much as you understand its boundaries. This article draws them.

🔑
One line — what it does well: filtering murmur cats for the echo, and telling cardiac from respiratory causes of laboured breathing. What it cannot do: replace the echo, or screen healthy cats wholesale. The numbers to remember are 100 and 270.

1. What it measures — the fragment a stretched heart sheds

Heart muscle releases a hormone called BNP when it is stretched — the more burdened the heart (rising pressures, a swelling atrium), the more it sheds. NT-proBNP is the fragment clipped off as that hormone is made; it survives longer and more stably in blood, which is why it is the piece we measure.

Its essence in one line: this test measures not "what disease the heart has" but "how much strain the heart is under right now." It cannot name a disease — but it answers "should we look harder at the heart?" with real accuracy.

2. Two versions — the one that gives a number, and the one that gives a colour

Quantitative (reference lab)SNAP point-of-care (in clinic)
ResultA number (pmol/L)Normal / abnormal, nothing between
Turnaround1–3 days10 minutes
UseInterpretable values, comparable over timeThe fast fork in an emergency
LimitSensitive to sample handling (section 5)Cannot show the grey zone

You can tell from the report — a number means quantitative; "Normal/Abnormal" alone means SNAP. Record which one it was: the two cannot be compared with each other.

3. The two jobs it does well — with the study numbers

Job ① The murmur cat's "echo or not yet?"

This is the role from the murmur article. In cats with murmurs it identified moderate-to-severe heart disease with 71% sensitivity and 92% specificity. The high specificity is the point: a high result makes the case for the echo fairly solid. A normal result supports "significant disease unlikely now" — but 71% sensitivity means roughly three in ten diseased cats can slip through, so the re-listening continues.

Job ② Sorting the breathless cat — this test's finest hour

Facing a dyspnoeic cat, the ER's most urgent question is "heart (oedema, effusion) or airways (asthma and kin)?" — because the treatments point in opposite directions. Here NT-proBNP posts its best numbers: at a cut-off around 265 pmol/L, sensitivity 90% and specificity 88%. Especially valuable at hours and clinics where an immediate echo is not an option.

A bonus: if there is pleural effusion, the tapped fluid itself can be tested. For "is this effusion cardiac?", it performs about as well as blood (bedside test: 100% sensitivity, 86% specificity). Worth asking about whenever a tap happens.

The job it does not do — screening healthy-looking cats wholesale

"No signs, but shall we check the heart by blood?" has weak footing. In cats with no murmur and no signs, a normal result does not guarantee health (early HCM mostly reads normal), and many abnormal results are not disease. This test earns its keep only in cats with a clue. The heart item of a wellness check is auscultation — and, when needed, the echo.

4. Reading the number — 100 and 270

Quantitative result (pmol/L)ReadingNext step
Under 100Normal — significant cardiac strain unlikelyIf there was a clue (murmur), keep the re-listening schedule
100 – 270Grey zone — strain may be presentEchocardiogram recommended. The vaguer the number, the more the answer is imaging
Above 270Strong signal — (in a breathless cat) the heart is the likely causeEcho — after stabilisation if it is an emergency

One caution: these cut-offs are foggy hills, not cliffs. 98 and 105 are effectively the same state, and labs vary. Do not live and die by numbers near a boundary — read the direction (vs last time) together with the other findings.

5. When the number lies — in both directions

SituationWhich way it fools youWorth knowing
High kidney valuesFalse riseThis fragment leaves through the kidneys. With creatinine high (roughly 2.8+), it can climb without cardiac strain — in a CKD cat, always interpret it beside the kidney values
HyperthyroidismA real rise — but indistinguishableThe whipped heart genuinely sheds more. The problem: the number cannot separate "thyroid heart" from true HCM — re-evaluate after thyroid treatment
Dehydration, systemic illness, hypertensionFalse/mixed riseA sick body strains the heart too; read mid-illness values conservatively
Sample handlingFalse fallThe molecule degrades in the tube — up to 30% lost within a day at room temperature in one report. Prompt processing and cold transport are conditions; a suspiciously low value can be a sample problem
Early diseaseReads normalMild B1 mostly sits in the normal range — why a normal result is not proof of "no HCM"

6. Why it cannot replace the echo — or track the disease

  • It knows neither the disease nor the stage — only that the number is high. Whether it is HCM or another heart disease, B1 or B2 (= whether clopidogrel starts) — all of that belongs to the echocardiogram. NT-proBNP is the bridge to the echo, never the terminus
  • It is weak as a progress tracker — even in the trial where rapamycin halted wall thickening (RAPACAT), NT-proBNP did not move. Whether treatment is working is read from the echo's three numbers and the breathing rate, not from this value
  • SNAP-normal is not a guarantee — the bedside version cannot show the grey zone. SNAP normal with a clue still standing (murmur, signs) should continue to the quantitative test or the echo

7. In practice — when to request it, how to record it

  • Three moments to ask for it — ① a murmur was found and the echo decision feels premature ② a breathless emergency needs the cardiac-or-respiratory fork fast ③ blood is being drawn anyway (e.g. CKD rechecks) in a cat with a cardiac clue on file
  • Record — date, value (or "SNAP"), and the kidney values from the same draw. An NT-proBNP filed without its creatinine cannot be interpreted later
  • Cost sense — far cheaper than the echo, but the accurate frame is not "the cheap substitute": it is the toll-gate fee for deciding whether to drive to the echo

8. What to ask your vet

  • "Was this the quantitative test or the SNAP? If quantitative, may I have the number?"
  • "Given this cat's kidney values, how should we read it?"
  • "(Grey zone) so the next step is the echo — when?"
  • "(Senior) did we check the thyroid alongside?"
  • "(At a tap) can we run this test on the pleural fluid too?"

9. Related

📌
One line — NT-proBNP is the toll gate that measures cardiac strain: under 100 leans safe, over 270 is a strong signal, and the space between belongs to the echo. And whatever the number, read it side by side with the kidneys.