Correspondence
NXY-059 for Acute Ischemic Stroke
N Engl J Med 2006; 354:2075-2076May 11, 2006
- Article
To the Editor:
In the article by Lees et al. (Feb. 9 issue)1 about the Stroke–Acute Ischemic NXY Treatment (SAINT I) trial, the abstract suggests significant reduction of hemorrhagic transformation after receipt of NXY-059, as determined by a post hoc analysis. These data are not associated with the primary or secondary outcomes of the trial, the groups were not powered for the appropriate statistical analysis, and there is a strong possibility of a plausible alternative explanation. Such a comparison is suitable for a future meta-analysis after data from multiple studies with NXY-059 are available and can be pooled together. Presenting post hoc data in the abstract of a single randomized trial should be avoided, especially when the data are based on 22 patients.
1 ReferencesVictor Serebruany, M.D., Ph.D.
Johns Hopkins University, Baltimore, MD 21204
heartdrug@aol.com 1
Lees KR ,Zivin JA ,Ashwood T , et al. NXY-059 for acute ischemic stroke. N Engl J Med 2006;354:588-600
Full Text | Web of Science | Medline
To the Editor:
Lees et al. have conflated statistical significance with clinical significance in extolling the efficacy of NXY-059 in acute ischemic stroke. The primary outcome measure of their trial was disability at 90 days, as assessed by the modified Rankin scale. From their Table 2 (after the pooling of scores 5 and 6 on the modified Rankin scale), the mean (±SD) modified Rankin score in the placebo group at 90 days was 2.84±1.75, as compared with 2.71±1.81 in the NXY-059 group. The observed difference (2.84−2.71=0.13) corresponds to an effect size (±pooled SD) of 0.07±0.13 in favor of NXY-059, quite modest by usual standards.1,2 A sober interpretation of the study findings provides little evidence for enthusiasm for NXY-059.
2 ReferencesJames A. Koziol, Ph.D.
Scripps Research Institute, La Jolla, CA 92037
koziol@scripps.edu 1
del Zoppo GJ . Stroke and neurovascular protection. N Engl J Med 2006;354:553-555
Full Text | Web of Science | Medline2
Cohen J. Statistical power analysis for the behavioral sciences. New York: Academic Press, 1969.
Author/Editor Response
We collected our data on hemorrhagic transformation prospectively. In criticizing our inclusion of the post hoc analysis of data about symptomatic hemorrhage, Serebruany disregards three points. First, previous experimental data demonstrate that free-radical–trapping agents combined with recombinant tissue plasminogen activator (rt-PA) can reduce hemorrhagic transformation.1,2 Second, our analysis included 105 events within a sample of 489 patients — a population similar to that of the National Institute of Neurological Disorders and Stroke rt-PA Stroke Trial3 — with significant differences among all 105 patients who had hemorrhagic transformation. Third, neuroprotection has long been sought as a possible means either to extend the time window or to enhance the safety of thrombolysis, or both, after acute ischemic stroke. We interpret our findings as supportive of the primary and secondary outcomes of our trial; that these outcomes are unrelated would seem to strengthen rather than weaken the inference of a biologic signal attributable to NXY-059. We have not seen any report that offers a plausible alternative explanation.
In contrast to the conclusions of Koziol, we regard the results of SAINT I as statistically robust and clinically meaningful. He overstates the standard deviation of the Rankin results by disregarding adjustment for baseline severity, as used in our primary analysis, leading to an erroneous interpretation of the magnitude of the benefit. Regardless of his error, modern medicine uses the concept of “number needed to treat” for a particular benefit. In the case of NXY-059, the number needed to treat is 22 for “cure” and less than 8 for improvement by one Rankin grade, with no associated harm. For comparison, consider the modestly enhanced survival after myocardial infarction resulting from the use of acute thrombolysis (number needed to treat, 33). Even though coronary thrombolysis carries some risk, that treatment is widely accepted and used. Moreover, even modest improvements to health in individual persons might be considered clinically meaningful when placed in the context of the eligible population. NXY-059 could be safely administered to a large proportion of the estimated 15 million persons who have stroke each year. We have reported a statistically significant finding of an effect that, if confirmed in SAINT II, would have far-reaching consequences for patients with stroke and clinicians.
3 ReferencesKennedy R. Lees, M.D.
Gardiner Institute, Glasgow G11 6NT, United Kingdom
k.r. lees@clinmed. gla. ac. uk Patrick Lyden, M.D.
University of California, San Diego, Stroke Center, San Diego, CA 92103Ashfaq Shuaib, M.D.
University of Alberta, Edmonton, AB G11 6NT, Canada1
Asahi M ,Asahi K ,Wang X ,Lo EH . Reduction of tissue plasminogen activator-induced hemorrhage and brain injury by free radical spin trapping after embolic focal cerebral ischemia in rats. J Cereb Blood Flow Metab 2000;20:452-457
CrossRef | Web of Science | Medline2
Lapchak PA ,Araujo DM ,Song D ,Wei J ,Purdy R ,Zivin JA . Effects of the spin trap agent disodium-[tert-butylimino)methyl]benzene-1,3-disulfonate N-oxide (generic NXY-059) on intracerebral hemorrhage in a rabbit large clot embolic stroke model: combination studies with tissue plasminogen activator. Stroke 2002;33:1665-1670
CrossRef | Web of Science | Medline3
The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group
Full Text | Web of Science | Medline
- Citing Articles (1)
Citing Articles
1
Hong Zhang, Zhan-Qiang Wang, Dan-Yang Zhao, Dong-Ming Zheng, Juan Feng, Li-Chun Song, Yumin Luo. (2011) AIF-mediated mitochondrial pathway is critical for the protective effect of diazoxide against SH-SY5Y cell apoptosis. Brain Research 1370, 89-98
CrossRef
