Cerebrolysin Is a Licensed Medicine Made of Pig Brain, and Its Own Label Says Nobody Can Measure It

01What it is

The Austrian label, the document a doctor prescribes from, says it in a sentence: 1 mL contains 215.2 mg of a proteolytic peptide fraction from pig brain protein, in water. Proteolytic means broken into pieces by enzymes. Pig brain protein means pig brain protein.

So this is an extract, in the sense that undid Epithalon here: the famous work used a cattle gland extract, while the thing people buy is a four amino acid chain you can draw. The American substance registry files this one as structurally diverse, its term for a substance with no definable structure.

Section 5.2, on pharmacokinetics, meaning what the body does with a drug once it is in, says a direct measurement cannot be carried out, because the fragments are similar or identical to ones the body already makes. The regulator's own document says the drug cannot be measured directly in the body.

02What the trials found

PubMed returns 667 records for cerebrolysin, 62 tagged as randomised controlled trials.

Ziganshina and colleagues published the seventh Cochrane version in 2023, pooling seven trials and 1,773 participants given Cerebrolysin, or a similar cattle brain peptide mixture, or a dummy, within 48 hours of a stroke. (PMID 37818733) Deaths from any cause came out at a risk ratio of 0.96, meaning the treated group died at 96 percent of the placebo rate, over a range the data are consistent with of 0.65 to 1.41, across 6 trials and 1,689 people, at Cochrane's own moderate confidence rating. Non fatal serious adverse events came out at 2.39, more than twice the placebo rate, range 1.10 to 5.23, across 3 trials and 1,335 people, the only pooled result in the review whose range stays above 1.

The line that tells you most is neither. Cochrane records that none of its included studies reported poor functional outcome, early death, quality of life, or time to return to work. Deaths and harms got pooled because they were all anyone reported.

Heiss and colleagues ran the largest trial, CASTA, in Stroke 2012: 1,070 patients, 529 on 30 mL into a vein daily for 10 days on top of aspirin, 541 on saline. Its confirmatory endpoint, a combined test of three stroke scales, showed no significant difference. (PMID 22282884) The 10.5 percent against 20.2 percent mortality at 90 days everybody quotes comes from the most severely affected subgroup, labelled in the paper as post hoc, meaning that group was picked out after the results were in.

The rehabilitation trials look strongest. CARS, from Muresanu and colleagues in Stroke 2016, randomised 208 patients to 30 mL into a vein daily for 21 days alongside a rehabilitation programme both arms got. On its primary measure, an arm function test at day 90, the effect size was 0.71, range 0.63 to 0.79, where 0.50 would be a coin flip. (PMID 26564102)

In vascular dementia, Cui and colleagues pooled six trials and 597 people, half of them on a dummy injection, and on the two cognitive scales they could combine, across three of those trials and 420 people, found a benefit of 0.36 standardised units, about a third of the spread of the patients' own scores. (PMID 31710397) Every result was graded very low quality, and the conclusion adds any benefit may be too small to matter to a patient.

In head injury I could not find a Cochrane review, searching the Cochrane Library and PubMed. What exists is the CAPTAIN series, whose first trial gave 46 patients 50 mL a day into a vein for 10 days, then two further courses of 10 mL a day for 10 days each. It missed its primary endpoint among everybody randomised and reached it only among those who finished as written. (PMID 31494820) The registry says it was terminated for poor recruitment.

A guideline reads the record differently. The 2021 European Academy of Neurology and European Federation of Neurorehabilitation Societies document recommends 30 mL a day into a vein for at least 10 days, weighing motor recovery where Cochrane pooled what the trials reported. (PMID 34152062)

03The mechanism papers have been withdrawn

Filter PubMed to retracted publications, search cerebrolysin, and ten records come back. The seven I opened are all mouse and rat work, retracted between 2025 and 2026, with no reason on the notices I opened: two mouse studies from Eliezer Masliah's group, a Rett syndrome mouse study, and three papers from Hari Shanker Sharma and Dafin Muresanu among them.

Masliah was removed as scientific director of the National Institute on Aging's neuroscience division in September 2024, after an investigation found falsification or fabrication in two of his publications. Alzforum reports that eight of the 132 papers in that dossier concern cerebrolysin. Muresanu co-authored three of the retracted papers and is first author of CARS. The trial is not impeached by the retractions: two different kinds of work. But a substantial part of the animal evidence explaining why this extract should work has been pulled from the record.

The money runs the same way. Cochrane says the manufacturer supported three of the multicentre stroke trials and judged two at high risk of bias for it. CARS was funded by EVER Neuro Pharma, three of its authors work in the company's clinical research department, and its statistician sits on its advisory board.

04How it compares

Most entries here end with a dose nobody can source. This one ends with a dose that is fully sourced and still does not transfer: the label's range is written for a clinic with a drip stand, and the range that circulates was set by what fits in a syringe.

05Where the dose came from

Here the answer is not "nowhere". The label prints a table: 10 to 30 mL a day for four weeks in dementia, 20 to 50 mL for 10 to 21 days in ischaemic stroke, the kind caused by a blocked vessel, 30 to 50 mL for the same span in bleeding stroke, 20 to 50 mL for 7 to 30 days in head injury. Most trials used a volume from inside it, though Skvortsova's low arm and CAPTAIN I's follow up courses gave 10 mL where their rows start at 20.

What circulates is a different animal. A vendor guide describes 5 to 10 mL a day, under the skin or into a muscle, five days on and two days off for four to eight weeks then a four week break, citing nothing for any of it. Reported here, not recommended. The five days on, two off is the label's own five applications a week, restated. The four week block is the label's four week course. The four week break is not: the label's gap between courses is two months.

The amount runs at or below the bottom of every row, its upper end of 10 mL landing exactly on the bottom of the dementia row. The reason it stops there looks mechanical rather than clinical, and the reading is mine: 5 mL is the most the label allows into a muscle, 10 mL the most allowed as a straight injection into a vein. Above that you need a drip stand.

Injection under the skin appears nowhere on the label, which lists muscle, vein and drip only. A PubMed search for cerebrolysin and subcutaneous returned seven records, every one a rodent study or one that does not name a human population.

06What could go wrong

The label rules the drug out in three situations: hypersensitivity to it, a seizure that will not stop, and severe kidney impairment. It warns that in epileptic conditions seizures may become more frequent, and among its very rare effects, fewer than one in ten thousand people, records isolated grand mal seizures, the kind that convulse the whole body, and allergic reactions up to a shock like state.

Notice how many entries carry the phrase "after too rapid administration": dizziness, palpitations or an irregular heartbeat, heat with sweating. The label's answer to several of its own adverse effects is to slow the drip, a hospital's answer, and my inference rather than the label's that someone injecting an ampoule at home does not have it.

07What nobody knows

Whether it works in stroke is open, and nothing above settles it. Cochrane pooled the outcomes that were reported and found nothing on death and a signal of harm; the rehabilitation trials measured arm function and found a large effect. Those are two questions, and the first could not answer the second because the trials did not report function in poolable form.

What a dose under the skin does. I found no human study by that route, and since the label says the drug cannot be measured in the body, that is not a gap a trial would easily close.

Whether the retractions change the clinical case. The withdrawn papers were about mechanism, not about whether patients improved, and I found no clinical result withdrawn. But the label's own account of why this should work leans on that literature, and I found no reassessment of it since.

Batch to batch consistency, which I did not examine at all.

08What is left

A licence is a document, and this one is real in a way almost nothing else here is, which also makes it the most useful thing a seller could be handed. Austria authorised it for patients. The buyers are healthy adults, and behind them sit two studies, from 1998 and 2000, with nothing larger or more recent that I could find.

The mixed trial record is ordinary. The stack around it is not. A substance with no definable structure. A label saying its own pharmacokinetics cannot be measured. An amount fixed by syringe limits. A route I could find no human study on. Ten withdrawn papers, the seven I opened all in mice and rats, under the story about why it should work. None of those layers can be checked by the person buying the ampoule.

If you have bought Cerebrolysin, check whether the ampoule carried an Austrian authorisation number, and whether the seller named the routes the label actually lists.

Frequently asked

Is Cerebrolysin an approved drug?+

In Austria, yes. It has been an authorised prescription and pharmacy only medicine there since 25 March 1996, under authorisation number 1-21380, for supportive treatment of cerebrovascular disorders, in particular Alzheimer type and vascular dementia, deficits following a stroke, and head injury, in adults and patients over 65. It is not approved in the United States. The FDA substance registry holds an identifier for it and an orphan drug code with the status Designated for frontotemporal dementia, which is a regulatory decision and not permission to sell the drug.

What is Cerebrolysin actually made of?+

Pig brain. The Austrian label states that 1 mL contains 215.2 mg of a proteolytic peptide fraction from pig brain protein in water, with sodium hydroxide and water for injection as the only other ingredients. Proteolytic means the protein has been broken into pieces by enzymes. Because it is an extract rather than a single molecule, there is no sequence for a certificate of analysis to confirm, and the American substance registry files it as structurally diverse, meaning it has no single definable structure.

Does Cerebrolysin work for stroke?+

The evidence points both ways, and which trial produced which answer matters. The 2023 Cochrane review pooled seven randomised trials and 1,773 participants given Cerebrolysin, or a similar cattle brain peptide mixture, or a dummy, within 48 hours of a stroke, found no difference in deaths from any cause, and found more than twice the placebo rate of non fatal serious adverse events. Cochrane records that none of its included studies reported poor functional outcome or quality of life at all. Separately, the CARS trial randomised 208 patients to 30 mL into a vein daily for 21 days alongside rehabilitation and reported a wide separation on an arm function test at day 90. CARS was funded by the manufacturer.

What doses were used, and what do people take?+

The Austrian prescribing label, which is a licensed dosing table and not a recommendation from this site, gives 10 to 30 mL daily for four weeks in dementia, 20 to 50 mL daily for 10 to 21 days in ischaemic stroke, the kind caused by a blocked vessel, 30 to 50 mL daily for 10 to 21 days in bleeding stroke, and 20 to 50 mL daily for 7 to 30 days in head injury. The label caps a straight injection into a muscle at 5 mL and into a vein at 10 mL, so every larger volume on that table is a slow drip in a clinical setting. What circulates outside a clinic, reported here as convention and not as advice, is 5 to 10 mL a day with no citation behind it, which runs at or below the bottom of every row of that table.

Has injecting Cerebrolysin under the skin been studied?+

Not in humans, as far as I could find. The Austrian label lists only injection into a muscle, injection into a vein and a slow drip, and the word for under the skin does not appear in it. I searched PubMed for cerebrolysin and subcutaneous and got seven records, every one a rodent study or one that does not identify a human population. The label also says the drug cannot be measured in the body directly, so there is no established way to say what a dose under the skin delivers.

What are the retracted Cerebrolysin papers?+

Filtering PubMed to retracted publications and searching cerebrolysin returns ten records. The seven I opened are all animal work, retracted between 2025 and 2026, and they include two mouse studies from Eliezer Masliah's group, a Rett syndrome mouse study, and three papers from Hari Shanker Sharma and Dafin Muresanu. The retraction notices I opened give no reason on their PubMed pages. No clinical result has been withdrawn that I found, and the human trials are a different kind of work from the withdrawn animal papers.

---

This content is for educational and informational purposes only and is not medical advice. Peptides discussed are research compounds and may not be approved for human use. Nothing here should be used to diagnose, treat, cure, or prevent any disease. Always consult a qualified healthcare professional before starting any peptide, supplement, or protocol. Individual responses vary. Do not self-administer compounds without proper medical supervision.