A twice-yearly RNA shot lowered blood pressure, but the pooled evidence is thin

Zilebesiran, an experimental drug that silences a gene in the blood-pressure control system and is dosed as an injection roughly twice a year, lowered office systolic pressure by 7.16 mmHg more than placebo, according to a meta-analysis that pooled four randomised trials [s1]. That is a real effect, but the pooled result carries a wide confidence interval, a signal that the effect may fade by six months, and more hyperkalaemia and injection-site reactions than placebo [s1].

The interest in a drug like this is its dosing. Most blood-pressure treatment depends on patients taking daily pills, and most patients, over time, do not take them reliably. Zilebesiran is a small interfering RNA — it switches off production of angiotensinogen, the starting protein of the renin-angiotensin-aldosterone system — and a single injection acts for months, which is why it has drawn attention as a possible answer to the adherence problem [s1].

What the meta-analysis pooled

The analysis, published in the European Journal of Clinical Pharmacology by an academic team independent of the drug's developer, followed a prespecified method: two reviewers extracted data, bias was assessed with the RoB 2.0 tool, and results were combined with a random-effects model [s1]. It included four studies covering 1,412 adults with mild to moderate hypertension [s1].

On the primary outcome, the placebo-adjusted reduction in office systolic pressure, the pooled mean difference was −7.16 mmHg, with a 95% confidence interval running from −13.63 to −0.70 [s1]. The interval matters as much as the point estimate: its upper bound sits just short of zero, meaning the combined trials do not rule out an effect close to none, even as their central estimate is a worthwhile drop. Ambulatory measures, which track pressure across a full day, were larger — a 24-hour systolic reduction of −9.92 mmHg (95% CI −18.74 to −1.10), a daytime reduction of −9.76 mmHg (−19.46 to −0.06), a nighttime reduction of −9.95 mmHg (−16.71 to −3.19) and a 24-hour diastolic reduction of −7.27 mmHg (−11.13 to −3.41) — but all with intervals wide enough to demand caution [s1]. The drug did what its mechanism predicts, cutting angiotensinogen by a standardised mean of −1.00 (95% CI −1.15 to −0.85) [s1].

Two findings temper the picture. In a subgroup analysis by follow-up length, the daytime systolic reduction was significantly greater at three months than at six months (P = 0.02), which the authors read as a possible waning of effect over time — an important question for a drug whose entire selling point is long duration [s1]. And overall adverse events were more common with zilebesiran than placebo, with an odds ratio of 1.46, including more hyperkalaemia and injection-site reactions, though serious adverse events did not differ significantly between groups [s1].

What the largest trial adds

The pooled numbers are dominated by a handful of trials, and the largest of them, KARDIA-2, is worth reading on its own because it tested the practical question of adding zilebesiran to an existing blood-pressure drug. It randomised 663 patients whose hypertension was inadequately controlled on a single agent to zilebesiran or placebo on top of that agent [s2]. The extra reduction in 24-hour systolic pressure at three months depended heavily on the background drug: −12.1 mmHg when added to a diuretic, −9.7 mmHg when added to a calcium-channel blocker, and −4.5 mmHg when added to an angiotensin-receptor blocker [s2]. That last figure is the telling one — when the background drug already blocks the same hormonal pathway, there is less left for zilebesiran to do.

KARDIA-2 also sharpened the safety signal the meta-analysis flagged. More patients on zilebesiran than placebo had hyperkalaemia (5.5% versus 1.8%), hypotension (4.3% versus 2.1%) and acute kidney failure (4.9% versus 1.5%), though the trial reported most episodes as mild and self-resolving [s2]. Those are the expected hazards of durably suppressing this hormone system, and a long-acting drug cannot be switched off quickly if they occur.

What is missing

The largest gap is the one that decides whether any blood-pressure drug is worth taking: outcomes. No trial here measured whether zilebesiran prevents heart attacks, strokes or deaths — only how far it moves the pressure reading [s1]. The pooled trials are small and short, and every one used a placebo comparator rather than an active blood-pressure drug, so they show that zilebesiran beats nothing, not that it matches or beats what patients already take; the meta-analysis authors say as much, calling for larger trials with active controls [s1]. The hint of a fading effect by six months, if real, would bear directly on how the drug is dosed.

What to watch

The outcomes trials now underway are what will settle whether the blood-pressure drop translates into fewer cardiovascular events; whether the six-month attenuation holds up in larger, longer studies; and how the hyperkalaemia and kidney signals look once the drug is used in broader populations.

This article describes pooled trial results for an investigational drug and is informational only. It is not medical advice and does not recommend any drug. Zilebesiran is not approved for use.

Sources

• [s1] Hussain M, Hassan M, Sohaib A, Ahmed T, Khan S, Hannan A, Dhedhi HAA, Naz A, Cheema AH, Ahmed F, Efficacy and safety of zilebesiran in adults with hypertension: a systematic review and meta-analysis of randomized controlled trials, European Journal of Clinical Pharmacology, 2026;82:252, published online 2026-09-16.
• [s2] Desai AS, et al., Add-On Treatment With Zilebesiran for Inadequately Controlled Hypertension: The KARDIA-2 Randomized Clinical Trial, JAMA, 2025;334:46-55, published online 2025-05-28.