Autoantibodies Against Oxidized Low-Density Lipoprotein in Cerebral Small Vessel Disease
Background and Purpose—Oxidized low-density lipoprotein (oxLDL) induces endothelial dysfunction and antibody formation. Because endothelial dysfunction is involved in cerebral small vessel disease (CSVD) (dilated Virchow Robin spaces, lacunar infarcts, and white matter lesions), oxLDL antibodies could play a role in CSVD pathogenesis. Therefore, we studied oxLDL antibodies in patients with high prevalence of CSVD: lacunar stroke patients and essential hypertensive patients.
Methods—A total of 158 lacunar stroke patients, 158 hypertensive patients, and 43 healthy controls were included. We determined levels of IgG and IgM against hypochlorite (HOCl) and malondialdehyde (MDA) oxLDL using ELISA (values in optical density).
Results—Patients with CSVD had higher levels of IgG-HOCl-oxLDL (0.77 versus 0.70; P<0.01), as well as lower levels of IgM-MDA-oxLDL (0.55 versus 0.65; P<0.05) than patients without such lesions. Higher IgG-HOCl-oxLDL levels were only independently associated with higher numbers of Virchow Robin spaces at the level of the basal ganglia (β=0.218; P<0.001).
Conclusions—An autoinflammatory process with lower levels of IgM antibodies and higher levels of IgG antibodies against oxLDL may be involved in CSVD.
Oxidized low-density lipoprotein (oxLDL) has an established role in the pathogenesis of atherosclerosis. It acts as a proinflammatory and proatherogenic compound by inducing auto-antibodies and endothelial dysfunction.1 Next to involvement in atherosclerosis, anti-oxLDL antibodies could also play a role in cerebral small vessel disease (CSVD).
In CSVD, endothelial dysfunction is considered to increase the permeability of the blood–brain barrier, with dilation of the perivascular (Virchow Robin) spaces, (a)symptomatic lacunar infarcts, and white matter lesions (WMLs) as sequelae.2 However, it is still unclear which mechanisms promote endothelial dysfunction in CSVD. Given their role in causing endothelial dysfunction in atherosclerosis, anti-oxLDL antibodies may contribute to the pathogenesis of CSVD as well.
We hypothesized that patients with CSVD have higher levels of antibodies against oxLDL than patients without such lesions or healthy controls. To test our hypothesis, we selected 2 patient groups with a high prevalence of manifestations of CSVD: first-ever lacunar stroke patients, as well as patients with essential hypertension.
We prospectively included 158 first-ever lacunar stroke patients presenting at the department of Neurology between May 2003 and December 2007. Lacunar stroke was defined as an acute stroke syndrome with a lesion on imaging compatible with the occlusion of a single perforating small artery (subcortical, demarcated, and a diameter of <15 mm on MRI).3 Furthermore, other possible causes (cardiac embolism, large vessel disease, or carotid stenosis) were excluded.3In addition, we included 158 consecutive hypertensive patients from the outpatient department of Internal Medicine. At inclusion, these patients were free of comorbidity.4 Forty-three patients who visited the neurological outpatient department with myogenic back pain or entrapment neuropathies served as “healthy” controls. They had no vascular or inflammatory disease, no hypertension, and no silent ischemic lesions on cerebral MRI.
Magnetic Resonance Imaging
Both standard T2-weighted and fluid-attenuated inversion-recovery sequences were used. Images were assessed by consensus by two experienced neurovascular researchers.5 In case of disagreement, the judgment of a third was decisive. We counted asymptomatic lacunar infarcts and used the Fazekas scale for WMLs.6 Extensive WMLs were defined as periventricular hyperintensities with extension into white matter, and/or beginning confluence of lesions or large confluent lesions in deep white matter. Silent CSVD was defined as the presence of one or more asymptomatic infarcts and/or extensive WMLs. Furthermore, we assessed another manifestation of CSVD, namely Virchow Robin spaces at 3 different levels with a predefined 3-point scale5: (1) <20; (2) between 20 and 50; and (3) >50.
To prevent confounding by acute phase responses in lacunar stroke patients, blood was sampled at or around 3 months after their event, in a stable clinical condition.
We determined IgG and IgM antibodies against 2 different forms of oxidized LDL (malondialdehyde [MDA]-modified and hypochlorite [HOCl]-oxidized LDL) according to methods we described earlier.7 Results are expressed as netto optical density (OD): for each patient, results of native LDL wells were subtracted from results of oxLDL wells. Interassay variability was <10%.
First, we compared autoantibody levels between patients with or without small vessel disease characteristics with Mann–Whitney tests. We also compared antibody levels between all 3 patient categories using Kruskal–Wallis test; whenever this test revealed significant differences, we used Mann–Whitney tests with correction for multiple testing (adjusted P<0.017) to evaluate differences between patients and controls. Finally, we determined independent relations between log-transformed IgG-oxLDL levels and MRI characteristics of CSVD using multivariate linear regression analyses. IgM-oxLDL values could not be transformed into normally distributed data. Unless indicated otherwise, we considered a probability value of <0.05 statistically significant.
See Table 1 for characteristics of patients and controls. Differences in antibody levels between patients with or without silent lesions are shown in Table 2, whereas results of the antibody measurements for the different patient groups separately are shown in the Figure.
The only significant independent predictor of the log-IgG-HOCl-oxLDL levels was the number of Virchow Robin spaces at basal ganglia level (B=0.033; 95% confidence interval, 0.017 to 0.049; β=0.218; P<0.001). Results were corrected for age, sex, hypertension, and patient category, as well as for coronary and peripheral artery disease.
In the present study, we demonstrated that IgM-MDA-oxLDL levels were lower in patients with CSVD, whereas IgG-HOCl-oxLDL were higher in these patients as compared to patients without CSVD. Furthermore, the IgG/IgM-MDA-oxLDL ratio was higher in patients with CSVD. Most interestingly, the only independent predictor for higher IgG-HOCl-oxLDL levels were higher numbers of dilated Virchow Robin spaces at the level of the basal ganglia. Our data suggest that an autoimmune reaction against oxLDL plays a role in CSVD.
It is unclear whether the immune response against oxLDL is harmful or protective. Our study would be compatible with a differentiated view on the role of oxLDL antibodies in CSVD, in which IgM-oxLDL antibodies seem to protect from hypertension related vascular damage in the brain (IgM-oxLDL levels are higher in hypertensive patients and patients without silent CSVD; results not corrected for age), whereas the higher levels of IgG-oxLDL in lacunar stroke patients and in patients with silent CSVD, as well as the relationship between higher IgG-oxLDL and dilated Virchow Robin spaces (results corrected for age, sex and vascular risk factors), suggest that these antibodies may be involved in the pathogenesis of small vessel damage.
In contrast to our findings with the IgG-HOCl-oxLDL, we consider the differences in IgM-HOCl-oxLDL values, although statistically significant, less relevant, because their concentration as reflected by OD values are fairly low. Low OD values are accompanied by high background noise, and therefore the clinical relevance of these antibodies, at least in CSVD, seems limited.
Our study has several limitations. First, our study design is cross-sectional, and therefore the associations we found could as well be an epiphenomenon of CSVD or stroke. Second, we confined patient selection to those with CSVD, because we aimed at the relation between CSVD and oxLDL antibodies; therefore, we cannot exclude that the anti-oxLDL response does not occur in patients with other stroke subtypes. Third, our MRI protocol did not include diffusion-weighted imaging to ascertain the presence of an acute, symptomatic lacunar infarct. However, most MR scans in lacunar stroke patients were made several weeks after the stroke, a time frame in which diffusion weighted imaging is not of additional value. Fourth, to further enhance the quality of our anti-oxLDL test, we could have used blocking to prevent aspecific antibody binding (although, in our hands, background binding was even higher; results not shown) or have used an commercially produced antibody for intraassay control. However, we suggest that only the use of a capture ELISA (using a commercially produced antibody as coating to capture oxLDL) could further lead to more specific results of the anti-oxLDL test. Notwithstanding these limitations, the strengths of our study remain that we examined a large group of well-characterized patients and that we excluded the effects of an acute-phase response.
In conclusion, our study provides evidence for an autoinflammatory process with oxLDL as a candidate antigen in CSVD. However, further research is needed; results from future longitudinal studies, as well as from basic research, should provide insight into the role that oxLDL antibodies play in the inflammation, which is suspected to contribute to CSVD.
Sources of Funding
This work was supported by The Netherlands Heart Foundation (2005B022 to R.R.), The Netherlands Thrombosis Foundation (2007-3 to I.K.), and the Novartis Foundation for Cardiovascular Excellence (003/07 to L.H.).
- Received June 5, 2010.
- Accepted August 12, 2010.
de Jong G, Kessels F, Lodder J. Two types of lacunar infarcts: further arguments from a study on prognosis. Stroke. 2002; 33: 2072–2076.
Henskens LH, Kroon AA, van Oostenbrugge RJ, Gronenschild EH, Fuss-Lejeune MM, Hofman PA, Lodder J, de Leeuw PW. Increased aortic pulse wave velocity is associated with silent cerebral small-vessel disease in hypertensive patients. Hypertension. 2008; 52: 1120–1126.
Slot MC, Theunissen R, van Paassen P, Damoiseaux JG, Cohen Tervaert JW. Anti-oxidized low-density lipoprotein antibodies in myeloperoxidase-positive vasculitis patients preferentially recognize hypochlorite-modified low density lipoproteins. Clin Exp Immunol. 2007; 149: 257–264.