IL-6 Variants in Ischemic Stroke.

Dear Editor,

I have read with interest a recent report by Kumar et al. [1] on the putative role of interleukin-6 (IL-6) in ischemic stroke. In their report, the authors refute an association of 2 promoter variants on the phenotype under study based on earlier case-control investigations. Data from this article, however, warrant a reappraisal of existing findings.

On one hand, numerous investigations were not retrieved. A quick search using the same databases as employed by the authors unveiled additional studies both for −572G/C [2,3] and for −174G/C [4,5,6] and around 10,000 additional genotypes that had been available at the time of the August 30, 2014, data freeze. I am not even counting the genome-wide association studies that were published from 2007 onward [7]. On the other hand, the results cited in the article appear to have been muddled. Thus, a study by Flex et al. [8] refers to an entirely different phenotype of peripheral artery occlusive disease (only 22 patients actually also had a history of stroke). Assuming that the authors had intended to refer to another study by the same author [9], we face the obvious overlap of cases and controls with a further investigation [10]. Similarly, overlap of cases and controls was ignored for the studies by Revilla et al. [11] and Chamorro et al. [12]. The number of cases and controls pooled is thus inflated. Another issue that has been overlooked is the sharp discrepancy in −174C allele frequencies in the studies by Yamada et al. [13] and Tong et al. [14]. No alleles have so far been identified with frequencies ranging from 0.22 to 0.76 in non-isolated Asian (or Caucasian) populations. The only reasonable explanation for this discordant observation is a muddling of major and minor alleles at some point, and therefore, the outlier study should have been dropped from the meta-analysis. On the whole, the present quantitative review is best reconducted to eliminate the above sources of bias and to provide an accurate estimate of IL-6 impact on the susceptibilty to ischemic stroke.

Disclosure Statement

None declared.