Similar to previous neuroprotective agents, AX200’s success in animal trials did not translate into positive results in clinical studies.
NEW ORLEANS—AX200, a proprietary form of granulocyte-colony stimulating factor (G-CSF), is no more effective than placebo at reducing the level of disability in patients with acute ischemic stroke, according to a study presented at the 2012 International Stroke Conference. AX200 is the latest drug candidate to fail a clinical study after showing promise in animal trials.
Investigators have not determined the cause of the drug’s failure, but are continuing to analyze the data. “Our results confirmed the principle problem in the translation of preclinical work into clinical stroke trials,” said E. Bernd Ringelstein, MD, Professor of Neurology at the University of Münster in Germany.
A Once Promising Candidate
In vitro and in vivo studies have demonstrated G-CSF’s multimodal action in the brain, which includes inhibiting apoptosis and stimulating neurogenesis, arteriogenesis, and neurite outgrowth. AX200 (filgrastim) is a recombinant human G-CSF approved in the United States for lowering the risk of infection in patients receiving chemotherapy and patients with severe chronic neutropenia.
A Phase IIa trial indicated that AX200 was well tolerated and safe. Based on this study, Dr. Ringelstein and his colleagues designed a randomized, double-blind, placebo-controlled multinational trial to demonstrate the efficacy of AX200 compared with placebo. The primary end point was modified Rankin scale score after three months, and the secondary end point was the NIH Stroke Scale (NIHSS) score at three months. The researchers also investigated mortality, adverse events, imaging, and hematology parameters.
Dr. Ringelstein and his colleagues enrolled 328 patients with acute ischemic stroke, mostly from Germany and Eastern Europe, at 78 sites. Eligible patients had had infarcts of at least 15 cm3 within the middle cerebral artery, NIHSS scores between 6 and 22, and were younger than 86. The investigators required patients to be treated within nine hours of symptom onset. Patients who had received t-PA before enrollment were accepted if they had met inclusion criteria after lysis therapy.
A total of 323 patients entered the intent-to-treat analysis, and 160 of them received 135 µg/kg of AX200 over 72 hours with an initial loading dose. The other patients received a placebo. The revision of all images was performed by a core imaging laboratory.
Patient demographics were similar in both arms of the study. Patients’ mean age was 69, and about half of participants were male. The mean baseline NIHSS score was approximately 12 for both groups.
Similar Results for AX200 and Placebo
“The G-CSF did exactly what it should do,” said Dr. Ringelstein. “It increased in the plasma, it strongly stimulated the production of white blood cells, and slightly decreased the number of platelets.”
After three months, however, the two groups’ modified Rankin scale scores were not statistically different. The mean score for the treatment group was 3.31, compared with 3.12 for the control group. The researchers also found a broad overlap of the confidence interval for the two groups.
In addition, the team found no significant difference between the two groups for the secondary end point. The mean NIHSS score for the treatment group was 8.88, compared with 8.45 for the control group.
When the investigators compared the primary and secondary end points for patients pretreated with t-PA and those not pretreated with t-PA, they found no difference and no trend. “Unlike erythropoietin, G-CSF showed no significant signs of adverse interaction with t-PA,” said Dr. Ringelstein.
The three-month death rate was in the expected range of about 20%. No statistically significant difference was noted between the treatment and control groups.
After adjusting for age, initial NIHSS score, and infarct volume, the team found no statistically significant difference and no trend in the Rankin shift analysis. In addition, no statistically significant difference was observed between the initial and final infarct volumes.
A Recurring Stumbling Block
“AX200 was a very promising drug with a comprehensive preclinical and clinical package,” said Dr. Ringelstein. But, similar to many other drug candidates for the treatment of stroke, the drug’s success in animal trials was not replicated in human trials.
The failure of AX200 may influence pharmaceutical companies to stop funding the development of drug candidates for stroke, Dr. Ringelstein told Neurology Reviews. “Maybe we should start at the very beginning again,” he commented. “Let us go back to the in vitro tests. Let us go back to basic science, firstly maybe not on the pharmaceutical companies’ level, but maybe on a more basic science level.”
—Erik Greb
To hear an audiocast related to this news article, please click here.