Intramyocardial haemorrhage named as a target, not yet as a treatment

Design

single-centre, non-randomised, placebo-controlled sequential-cohort phase IIa study with blinded cardiac MRI core laboratory (SHIELD-MI)

Population

25 patients given dexrazoxane before primary PCI and 25 matched placebo comparators selected from 78 placebo-treated patients with STEMI

Primary outcome

left ventricular ejection fraction and intramyocardial haemorrhage volume on cardiac MRI at 48-72 hours

Effect

LVEF 39.8+/-7.7% vs 34.7+/-10.1% (P=0.048); haemorrhage 2.0+/-3.4% vs 6.3+/-6.0% of LV (P=0.004)

SHIELD-MI gave intravenous dexrazoxane 250 mg before primary PCI and again at 4, 8 and 12 hours to 25 patients with ST-elevation myocardial infarction. Twenty-five comparators were chosen from 78 previously placebo-treated patients and matched on total ischaemic time, culprit territory and pre-PCI occlusion. Cardiac MRI at 48 to 72 hours measured left ventricular ejection fraction and intramyocardial haemorrhage volume, read by a blinded core laboratory.

Ejection fraction was higher with dexrazoxane (39.8 plus or minus 7.7% versus 34.7 plus or minus 10.1%, P=0.048) and haemorrhage volume lower (2.0 plus or minus 3.4% of the left ventricle versus 6.3 plus or minus 6.0%, P=0.004). Infarct size was smaller (29.1% versus 43.8% of LV, P=0.002), and haemorrhagic infarction occurred in 24% versus 64%. No drug-related serious adverse events were seen.

Intramyocardial haemorrhage after reperfusion is common, predicts adverse remodelling and has had no treatment aimed at it. Dexrazoxane is an established iron-chelating cardioprotectant from oncology, so the mechanism is plausible and the drug is familiar.

The design is the limit. This was single-centre, non-randomised and sequential-cohort, with comparators selected from a larger placebo pool, which is exactly the arrangement in which unmeasured differences favour the treated group. Fifty patients and a P of 0.048 on the gating endpoint leave the result fragile. Nothing here justifies using dexrazoxane in STEMI outside a trial; what it justifies is running one.

• Do not add dexrazoxane to primary PCI protocols on this evidence
• Where post-STEMI cardiac MRI is available, note that haemorrhage is reported separately from microvascular obstruction and predicts worse remodelling
• Total ischaemic time remains the modifiable variable that actually changes infarct size today
• Read effect sizes from matched sequential cohorts as hypothesis-generating, whatever the P value
• Flag suitable STEMI patients for enrolment if a randomised trial of this approach opens locally

The statistics, in plain English

With 25 patients per group, these differences are large but unstable: the ejection fraction result at P=0.048 sits just inside conventional significance and would not survive much loss of data. Because comparators were selected from a larger placebo pool rather than randomised, any characteristic not used in the matching can explain part of the difference, and matching on three variables cannot rule that out. The consistency across three related measures, haemorrhage volume, infarct size and ejection fraction, is what makes the signal interesting; it is not what makes it reliable.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.