CASTA's published abstract reports a neutral confirmatory endpoint. Its favorable severe-stroke signal was post hoc and requires confirmation. This record is deliberately limited to accessible primary abstracts, indexed publisher first-page text, and the official trial registry, not a fresh full-paper or patient-data analysis.[1,2]
The Study Library keeps individual trials separate from compound overviews. The question here is what these sources support about CASTA, and where they do not support a reliable conclusion.
Study at a glance
| Field | Evidence available for this record |
|---|---|
| Trial | CASTA; NCT00868283 |
| Comparison | Cerebrolysin versus saline placebo, with aspirin as background therapy |
| Setting and timing | Acute ischemic hemispheric stroke; randomization within 12 hours of symptom onset |
| Follow-up | Ten-day assigned-treatment period; follow-up through day 90 |
| Reported allocation | The results abstract reports 1,070 participants: 529 assigned active treatment and 541 placebo |
| Primary test | Combined global directional test of modified Rankin Scale, Barthel Index, and NIH Stroke Scale |
| Main result | No statistically significant between-group difference on the confirmatory endpoint |
| Registry status | Completed; no results posted; last update posted December 15, 2023 |
| Review boundary | Abstract-and-registry evidence; no independently verified detailed safety table, flow diagram, statistical analysis plan, or supplement |
The comparison identifies the research question, not a treatment protocol for readers. The guide to reading a research study explains why a paper's question, analysis and conclusion need to be considered together.
What the primary result means
The three measures address disability, daily functioning and neurological deficit. The 2009 design abstract describes their combination into one global test at day 90. That is more specific than simply naming three scales.[5]
A combined test asks a defined statistical question. It does not let a reader choose whichever component later appears most favorable. A neutral result is not proof that two interventions are identical, but it does not establish the proposed overall advantage. Nor does an encouraging biological rationale supply the missing clinical evidence.
For related interpretation, see negative and null study results. This record does not reconstruct the global statistic, its exact P value, or the missing-data procedure from inherited draft text.
Why the severe-stroke signal remains exploratory
The results abstract identifies analyses in participants with baseline NIH Stroke Scale scores above 12 as post hoc. Its conclusion calls for a further trial.[1]
Selecting a subgroup after the main analysis changes the evidentiary role of a result. A promising observation can motivate a new research question without confirming that the treatment works for that subgroup. An apparent difference within one subgroup also does not, by itself, prove a treatment-by-subgroup interaction.
Exact exploratory effect estimates are omitted here because the complete intervals, analysis context and underlying tables were not verified. No confirmed survival benefit, responder benefit, or overall mortality conclusion is inferred from the abstract's selected subgroup account.
Registration came after recruitment began
The official record reports a September 2006 study start. Its earliest submission was March 23, 2009, with first posting estimated March 24, 2009. Registration therefore occurred after the reported start, not before recruitment.[3,4]
The earliest and current records were read separately. Their main chronology is:
| Record | What the displayed fields establish |
|---|---|
| Earliest version, submitted March 23, 2009 | Recruiting; start September 2006; estimated primary completion August 2009 and study completion November 2009 |
| Earliest planning counts | Top-line enrollment 1,056 estimated; narrative approximately 1,060 planned |
| Current completion fields | Primary completion October 2010; study completion February 2011 |
| Current enrollment fields | Top-line enrollment 1,071 actual; narrative 1,070 |
| Current primary outcomes | Modified Rankin Scale, Barthel Index and NIH Stroke Scale, each at 90 days |
| Current reporting state | Completed, but no structured results submission posted |
Planning counts are not final randomized or analyzed denominators. The current one-person enrollment mismatch remains unresolved; it is not silently assigned to a participant-flow category.
The 2009 design report was published while the trial was underway. It is useful evidence of the described design, but it cannot be treated as a public protocol frozen before the reported study start.[5]
Only the earliest and current registry content were examined, not a complete comparison of every intermediate version. The registry's listed scales do not themselves specify the publication's global decision rule. Agreement in scale names does not establish that every analysis choice was prespecified, and late registration alone does not prove outcome switching. The trial-registration guide explains that distinction.
What cannot be established about missing data or harms
The accessible results abstract says adverse events were recorded, but it does not supply a usable detailed harm table or complete participant disposition.[1] The current registry has no posted results tables.[3]
Consequently, this record does not reproduce arm-specific serious-adverse-event counts, event-versus-person counts, withdrawals, completers, per-protocol exclusions, or purported safety and efficacy denominators from an older draft. It also does not conclude that safety was equivalent or that no serious harm occurred.
The number assigned to an arm cannot automatically be reused as the denominator for every outcome. Follow-up, exclusions and missing observations matter, and the method used to handle them can affect an estimate. Without the relevant tables and methods, uncertainty should remain visible rather than being filled with assumptions. The analysis-set guide explains these distinctions.
This is a limit on the current review, not a claim that the full paper contains no additional information. Verified full-paper access could support a separately documented update.
Sponsor and disclosure context
The current registry names Ever Neuro Pharma GmbH as sponsor. The accessible 2009 design-publication page identifies funding from EBEWE Neuro Pharma GmbH and gives a no-conflict declaration for that report.[3,5]
Those statements concern their respective sources. They do not establish the full sponsor responsibilities or individual author disclosures in the unread 2012 article. This record therefore does not repeat unverified claims about data collection, statistical independence, committee roles or author payments.
Funding is relevant context, not a reason to accept or reject a result automatically. Transparent methods, endpoint hierarchy, completeness of reporting and access to evidence remain the more direct questions.
The tested mixture is not a generic peptide claim
Indexed first-page text from the primary publication describes Cerebrolysin as a porcine-brain-derived preparation containing low-molecular-weight peptides and free amino acids.[2] The evidence concerns that trial preparation and comparison.
A complex mixture is not one isolated peptide. A similar product name, concentration claim or analytical certificate does not establish equivalence to the trial material. CASTA cannot authenticate a research vial, another hydrolysate, a reformulation or an unspecified peptide blend. The neurobiology research hub provides broader context without collapsing distinct test articles into one category.
This record is not a current treatment recommendation or an approval determination. It does not compare the intervention with every contemporary stroke-care option, establish long-term benefit, or provide preparation or administration instructions.
Frequently asked questions
Why lead with a neutral endpoint?
The primary test anchors the study's intended main question. Highlighting a favorable later analysis without that context would change the reader's understanding of the evidence.
Does a post hoc finding have no value?
It can generate a hypothesis. Its value is different from a prospectively defined, adequately tested confirmatory claim.
Which enrollment number should be used?
Use the number and definition attached to the particular source or analysis. Do not merge estimated recruitment targets, enrollment fields and analyzed populations into one supposedly universal denominator.
Does late registration prove misconduct or outcome switching?
No. It limits what a public record can establish about planning before enrollment. Determining changes would require a sufficiently complete, dated comparison of the relevant documents and versions.
Why omit detailed adverse-event and dropout numbers?
They were not verified from the necessary full-paper tables in this review. An incomplete review should not present inherited figures as newly checked findings.
Does no posted registry result mean no paper exists?
No. A journal publication and a structured registry results submission are different reporting objects. One does not automatically replace the other.
Could this trial validate a differently sourced mixture?
No. Material identity and clinical evidence must match the actual test article; names alone cannot bridge that gap.
Primary and official sources
- Results abstract and bibliographic record: Heiss et al., Stroke 2012; PMID 22282884
- Primary publisher article: DOI 10.1161/STROKEAHA.111.628537; indexed first-page text only in this review
- ClinicalTrials.gov: current NCT00868283 record
- ClinicalTrials.gov: earliest NCT00868283 version, submitted March 23, 2009
- Design-report abstract and publication notes: Hong et al., 2009; DOI 10.1111/j.1747-4949.2009.00340.x
Review scope and corrections
Editorial evidence review: September 3, 2026, in America/Edmonton and UTC. Serum Research is responsible for this source-based educational summary. The review used the accessible abstracts, indexed primary publisher first-page text, and the current and earliest official registry records. It did not verify the complete results article, supplement, all intermediate registry versions, or current correction/retraction relationships. No independent clinical or statistical specialist review is claimed.
Return to the Research Library. To report a source discrepancy or correction, use Contact Us.