Cerebrolysin and Disorders of Consciousness: A Pilot Trial Against a Cochrane Verdict

Abstract

A pilot trial in Seoul is giving Cerebrolysin — a peptide preparation derived from pig brain — or saline to twelve patients who have been in a vegetative or minimally conscious state for more than four weeks after a haemorrhagic stroke.

Twelve participants is very small. The design around them is not: quadruple masking, and PET imaging alongside a behavioural scale as co-primary outcomes.

The reason to write about it now is the evidence it sits on. A Cochrane review published in 2023 concluded that Cerebrolysin probably has no beneficial effect on death after ischaemic stroke, and identified a possible increase in non-fatal serious adverse events. This note sets that record beside the trial, and explains why the trial is nonetheless asking a question Cochrane did not answer.

1.What Cerebrolysin Is, and Where

Cerebrolysin is a mixture of low-molecular-weight peptides and amino acids produced by enzymatic breakdown of purified brain protein. It is given intravenously, it is a prescription medicine rather than a supplement, and it is licensed in a number of countries in Central and Eastern Europe, Asia and the former Soviet states.

It is not authorised across much of the European Union, and nothing in this article should be read as suggesting anyone seek it. We cover it because it occupies an unusual position: a neuroprotective product used at scale in clinical practice for decades, with an evidence base that has repeatedly failed to support it.

2.What Cochrane Found

Liliya Ziganshina and colleagues have maintained a Cochrane review of Cerebrolysin in acute ischaemic stroke through several updates; the 2023 version included seven randomised trials with 1,773 participants, six of them contributing 1,689 participants to the pooled analyses.

On all-cause death, the risk ratio was 0.96 with a confidence interval from 0.65 to 1.41 — no effect, at moderate certainty. On the total number of people experiencing serious adverse events, 1.16 with an interval from 0.81 to 1.66 — again nothing, again moderate certainty.

The third number is the one that matters. For non-fatal serious adverse events the risk ratio was 2.39, with a confidence interval from 1.10 to 5.23, at moderate certainty. That interval excludes 1. The review's own summary describes it as a potential increase.

The authors' conclusion states that Cerebrolysin probably has no beneficial effect on preventing all-cause death in acute ischaemic stroke, and that the evidence indicates a potential increase in non-fatal serious adverse events.

The review also documents where the underlying trials came from. The manufacturer supported three multicentre studies — wholly, or by supplying the drug and placebo, randomisation codes, research grants or statisticians. Two studies were judged at high risk of bias on that basis. Five carried unclear risk from losses to follow-up between 16 and 29 per cent, retrospective protocol registration, or unavailable protocols.

3.The Same Question, Answered Differently

In March 2026, Abdullah Afridi and colleagues published a systematic review and meta-analysis in Brain and Behavior of Cerebrolysin as an adjunct to mechanical thrombectomy in acute ischaemic stroke.

It pooled three observational studies with 294 patients and reported that good functional outcome was more common with Cerebrolysin, at a risk ratio of 1.56 with an interval from 1.25 to 1.93. Symptomatic intracerebral haemorrhage was less common, at 0.12 with an interval from 0.03 to 0.48. Mortality was reported as 64 per cent lower, at 0.36 with an interval from 0.18 to 0.68.

Set that against the randomised evidence: three observational studies with 294 patients produce a large mortality benefit; six randomised trials with 1,689 participants produce a risk ratio of 0.96.

This is not a puzzle. It is the most reliably reproduced finding in clinical epidemiology. Observational comparisons of who received an optional adjunct treatment and who did not are shaped by who was well enough to receive it, and the effect of that selection routinely exceeds the effect of any drug.

A separate assessment published in Medicine in 2024 reviewed 14 documents — eight systematic reviews and six pharmacoeconomic studies — and reached a favourable overall verdict on effectiveness and cost-effectiveness. Its searches drew heavily on regional databases. A literature can be large, consistent and favourable without being independent of the interest that produced it.

4.Why the Seoul Trial Is Still Worth Watching

Everything above concerns acute ischaemic stroke. The trial at Konkuk University Medical Center concerns something else: prolonged disorders of consciousness following haemorrhagic stroke, at least four weeks after onset, in patients assessed as vegetative or minimally conscious on the Coma Recovery Scale–Revised.

That is a different population, a different injury, a different time window and a different outcome. Cochrane's conclusion does not transfer to it, and it would be sloppy of us to pretend otherwise.

The design is also better than its size suggests. Twelve participants receive 30 ml of Cerebrolysin in saline or saline alone, intravenously, once daily from day 4 to day 17, under quadruple masking — participants, investigators, care providers and outcome assessors all blinded. Ever Neuro Pharma, the manufacturer, is listed as a collaborator, which is the same relationship Cochrane flagged and which is at least declared here in advance.

The co-primary outcomes are the Coma Recovery Scale–Revised and positron emission tomography, each at two days and at seventeen days. Pairing a behavioural scale with a metabolic image is the right instinct for this population, and it connects to a problem we take up separately in our work on covert consciousness: in a patient who cannot report anything, a bedside scale measures what the examiner can elicit, not what the patient is experiencing.

With twelve participants, no result will be conclusive about efficacy. What a pilot of this design can produce is a defensible estimate of whether the two measures move together, which is what a larger trial would need in order to be designed at all.

5.How We Would Read It

Do the scale and the imaging agree? If the Coma Recovery Scale improves and cerebral metabolism does not, in twelve patients under quadruple masking, the most likely explanation is the ordinary fluctuation of a condition that fluctuates.

Are serious adverse events reported in full? Cochrane's one signal against this compound concerns non-fatal serious adverse events. A pilot trial cannot confirm or refute a risk ratio of 2.39, but it can report its events completely, and failing to do so would be the most informative omission available.

And is it framed as a pilot when it publishes? The registry says pilot study. Twelve patients supports a feasibility claim and an effect-size estimate for planning. It does not support a sentence beginning "Cerebrolysin improves".

Editorial Comment

Cerebrolysin is a prescription medicine, not a supplement. MindHeaven® has no product resembling it, no interest in it, and nothing in this article is guidance about its use or about the care of anyone with a disorder of consciousness. Those decisions belong to a treating neurologist and to a family, and nothing written here should enter that conversation.

We publish it because it is the clearest case in this section of a pattern that also runs through the supplement category: a product with decades of clinical use, a large and favourable regional literature, an approving health-technology assessment — and, underneath all of it, six randomised trials producing a risk ratio of 0.96 for death and a signal of increased harm.

Longevity of use is not evidence. Volume of literature is not evidence. When the randomised evidence and the observational evidence disagree by this much, the randomised evidence is not one opinion among several.

We should also record what we could not read. The neuroimaging study of emerged minimally conscious states that we had filed as a source for this topic, and a review of cell-penetrating peptides, are both closed access with no open copy. The Cochrane review of Cerebrolysin for vascular dementia is likewise closed. We have written only from what we could read, and this article would be better if those three were open.