Cerebrolysin in patients with acute ischemic stroke in Asia: results of a double-blind, placebo-controlled randomized trial
Heiss WD, Brainin M, Bornstein NM, Tuomilehto J, Hong Z
Stroke (2012) · n=1,070
CASTA is the largest Cerebrolysin acute-stroke RCT ever conducted — 1,070 patients across Asia — and the primary endpoint was negative. The trial is the structural reason every honest reading of the Cerebrolysin literature has to address how to reconcile a large negative pivotal-scale trial with a positive manufacturer-aligned meta-analysis.
CASTA — Cerebrolysin Acute Stroke Treatment in Asia — is the largest randomized controlled trial of Cerebrolysin ever conducted, published in Stroke in 2012 (NCT00868283). The trial randomized 1,070 patients (529 Cerebrolysin, 541 placebo) presenting with acute hemispheric ischemic stroke within 12 hours of symptom onset to either 30 mL/day intravenous Cerebrolysin or matching placebo for 10 days, alongside aspirin as background standard care. The trial was conducted across multiple centers in Asia (China and other regional centers). The primary outcome was a confirmatory combined global directional endpoint over the modified Rankin Scale, Barthel Index, and NIHSS at day 90 — a Wei-Lachin pooled statistic that the investigator network has used across the broader program as an alternative to a single-scale primary endpoint. The combined global outcome was not significantly different from placebo; the trial's pre-specified primary analysis was negative.
The post-hoc subgroup analysis that has propagated through the secondary literature concerns severely affected patients (NIHSS >12 at baseline), where 90-day cumulative mortality was 10.5% on Cerebrolysin versus 20.2% on placebo. The authors framed this signal as hypothesis-generating rather than confirmatory, and the Cerebrolysin and Recovery After Stroke (CARS-1) trial was subsequently designed in part to test the severely-affected-patient signal prospectively. Tolerability across the trial was acceptable, with the safety profile dominated by stroke-context complications.
This trial is the load-bearing data point in any honest reading of the Cerebrolysin acute-stroke literature: the largest patient-number commitment the manufacturer's investigator network has produced, with a pre-specified primary endpoint that did not separate from placebo. Every favorable manufacturer-aligned meta-analysis (Bornstein et al. 2018) and every skeptical independent review (Cochrane 2023) routes through how to weight CASTA's headline result against the rest of the trial body. The Cerebrolysin page treats both readings.
CASTA's negative primary endpoint is the structural fact that the favorable secondary readings have to reconcile. The trial's use of a combined global directional endpoint over three scales is a design choice that the program defends as more rehabilitation-relevant than a single-scale endpoint, but it is not the conventional regulatory primary endpoint convention in stroke (modified Rankin Scale at 90 days), and the choice opens the trial to the structural critique that an endpoint built from three scales gives more degrees of freedom for post-hoc reframing. The severely-affected-patient subgroup signal (90-day mortality 10.5% vs 20.2% in NIHSS >12) is striking in absolute terms but is a post-hoc analysis with the corresponding multiple-comparison and population-selection caveats; the prospective CARS-2 trial in severely affected patients did not replicate the mortality signal at the same magnitude. The trial was sponsored by EVER Neuro Pharma. Conduct was across Asian centers — the patient population, clinical context, and standard-care background may not generalize directly to Western stroke-unit practice. The 10-day treatment duration is short relative to the rehabilitation-context dosing in CARS-1 (21 days). The failed peptide trials archive covers CASTA's place in the broader landscape of stroke-recovery agents whose pivotal-scale trials missed the primary endpoint after promising Phase II signals. This is the central trial through which the Cerebrolysin literature has to be read.
07·Member discussion
No member discussion yet.
Member-only conversation lives here — cycle notes, practitioner commentary, pattern-matching. Be the first paying member to start the thread.