Smoking is associated with increased lesion volumes and brain atrophy in multiple sclerosis

Considerable evidence indicates that both genetic and environmental factors are involved in multiple sclerosis (MS) susceptibility and disease progression.¹ Cigarette smoking is one of the most compelling environmental risk factors linked to the development and worsening of MS. The other environmental risk factors implicated include sun exposure/vitamin D deficiency and Epstein–Barr virus infections.^2,3

Recent studies have shown a relationship between cigarette smoking and MS,^1,4 although some earlier studies did not find evidence for a link.^5–7 Subjects who smoked 20 to 40 cigarettes a day had twice the risk of developing MS compared with nonsmokers,⁴ and smoking may interact with the progression of MS.^4,8–10 Smoking was recently shown to be a risk factor for conversion from clinically isolated syndromes (CIS) to clinically definite MS; the interval to the first relapse occurred sooner in smokers and the hazard ratio for conversion over 3 years for smokers was 1.8 compared with nonsmokers.¹¹ Smoking worsens the prognosis for progressive disease,¹² and the hazard ratio of secondary progression was 3.6 for ever-smokers compared with never-smokers.⁸ Cigarette smoking can also exacerbate MS, both chronically and acutely.^2,9

MRI provides the best available surrogate markers for monitoring the MS disease process and for assessing the effects of treatments and can therefore complement the currently available epidemiologic evidence for the role of smoking in MS disease progression. Numerous studies have shown MRI measures to be generally more sensitive than clinical measures such as the Kurtzke Expanded Disability Status Scale (EDSS) and the Multiple Sclerosis Functional Composite because it can provide evidence for subclinical disease activity. A diverse array of MRI metrics capable assessing pathologic processes ranging from transient blood–brain barrier breakdown and lesion dynamics to more irreversible atrophy is available. The mechanisms that relate cigarette smoking to MS risk or progression are not well understood, and the extent of anatomic brain damage associated with smoking in patients with MS has not been characterized. The goal of this study is to compare the MRI characteristics of smokers and nonsmokers with MS to better understand the role of cigarette smoking in MS and further identify underlying mechanisms that relate to MS disease progression.

METHODS

RESULTS

Lesion-related MRI measures.

The mean values of the key lesion-related (CE lesion number, T2 LV, T1 LV) and central atrophy (LVV) MRI measures are summarized in the figure. The mean number of CE lesions for the ever-smoker group was 1.2 ± 4.1, compared with 0.72 ± 2.5 in the never-smoker group (figure, A). Poisson regression analysis indicated that the ever-smoker status was associated with greater number of CE lesions (Wald χ² for smoking status variable = 17.0, p < 0.001). In regression analysis, the association between CE LV with smoking status had a positive slope of 0.90 (SE = 0.44) and p = 0.043. These results indicate that ever-smokers are at higher risk for developing more numerous CE lesions.

• Download figure
• Open in new tab
• Download powerpoint

Figure Mean values for the number of contrast-enhancing lesions, T2 lesion volume, T1 lesion volume, and lateral ventricle volume

Summary of the mean values for the number of contrast-enhancing lesions (A), T2 lesion volume (B), T1 lesion volume (C), and lateral ventricle volume (D) in the never-smoker, ever-smoker, and active smoker groups. The error bars are standard errors.

The regression results summarized in table 2 demonstrate that smoking is associated with increased T2 LV and T1 LV.

DISCUSSION

In patients with MS, smoking was associated with higher increased lesion burden and greater brain atrophy. Our results indicate that a wide range of quantitative brain MRI markers is affected by smoking in patients with MS.

MS ever-smokers in our study had higher EDSS scores than never-smokers (median EDSS score 3.0 in ever-smokers vs 2.5 in never-smokers, p ≤ 0.001). This finding supports previous reports showing that cigarette smoking increases the risk for disease progression.^8,12

In our cohort, we observed an association between smoking and an increased number of CE lesions. Nicotine, a major component of cigarettes, has been shown to increase microvascular blood flow on the brain and to raise the influx of permeable solutes across the blood–brain barrier in rats.²² This is important because leakage of the blood–brain barrier has been suggested as a factor in initiating the development of MS, and may be relevant to the recent report that cigarette smoking increases the risk of converting from CIS to clinically definite MS.¹¹ Cigarette smoking causes a transient worsening of motor functions in patients with MS compared with healthy controls.²³

We discuss some limitations of our study. Some of the shortcomings are related to the nature of self-reported questionnaire used. For example, we have not been able to consider secondhand smoke. Children exposed to parental smoking were found to have a higher MS risk: the duration of exposure affected the level of risk²⁴ and the incidence of MS increased with the cumulative exposure to smoking. Another potential criticism is our definition of the ever-smoker group. Most epidemiologic studies, e.g., National Health and Nutrition Examination Survey III, require exposure to 100 or more cigarettes,²⁵ which may involve an exposure period of 3 to 12 months of active smoking, depending on individual-specific changes in smoking patterns in response to the evolution of nicotine dependence. Given the reduced social acceptance and stigma associated with smoking in the United States, we expected that subjects would harbor some reservation regarding smoking behaviors of a limited nature, and we selected the cutoff of 6 months based on reasoning that subjects would be likely to recall and more willingly provide information regarding a 6-month period of cigarette smoking. A different, less stringent definition of past smoking could in theory influence the distribution of sample size between the ever-smoker and never-smoker groups.

Patients with MS exhibit a higher frequency of behaviors including cigarette smoking, alcohol abuse/dependence, and lower levels of physical activity that are known to adversely affect health outcomes.²⁶ Smoking can also be associated with other comorbidities and with other environmental risk factors for MS, such as increased anti-Epstein–Barr virus antibodies.^26,27 Our exclusion criteria consisted of conditions that may influence quantitative MRI analysis, including cerebrovascular disease and positive history of alcohol abuse. However, we did not obtain data on other comorbidities, such as obesity, hypertriglyceridemia, and hypertension.

African-Americans tend to have a more severe MS disease process, with rapid progression of disability and poor response to first-line disease-modifying therapies.²⁸ There is increasing evidence that smoking patterns differ between African-Americans and white Americans.²⁹ Our study did not have sufficient African-Americans and was not powered to address the contributions of racial differences in smoking behavior.

We also did not observe associations between the years of smoking or pack-years smoked. The exact reasons for the lack of association are not clear but might be related to the self-reported nature of the questionnaire. However, several other studies of smoking in MS have reported associations with the ever-smoker vs never-smoker status,^8,11,12 but their results with the continuous variables of duration of smoking and pack-years of smoking have not been emphasized to the same extent. Despite these limitations, the advantage of the panel of MRI measures we used is that they can assess both transient pathologic processes such as blood–brain barrier breakdown and long-term, relatively irreversible processes such as cumulative neurodegeneration.

The biological basis of the potential link between smoking and MS has not yet been fully elucidated. In addition to nicotine, cigarette smoke contains hundreds of potentially toxic components, including tar, carbon monoxide, and polycyclic aromatic hydrocarbons.⁹ Two phases of cigarette smoke exist: a tar or particulate phase and a gaseous phase, both of which contain extremely high concentrations of free radicals. Several hypotheses based on biological mechanisms link cigarette smoking to MS and to other autoimmune diseases such as rheumatoid arthritis and systemic lupus erythematosus.^9,30,31 The most frequent suggested mechanisms include possible vascular effects, effects on the immune system, increased production of nitric oxide, increased frequency of respiratory infections, and a potential direct neurotoxic effect of cyanides and other components of cigarette smoking.⁹

Cigarette smoke also affects the influx and activation of neutrophils, macrophages, and monocytes,³² and in addition, it increases the expression of the activation marker Fas (CD95) on B and CD4 T-lymphocyte cell surfaces.³³ Smoking is also associated with increases in C-reactive protein, interleukin 6, and urinary thromboxane metabolites, which are important markers of inflammation in autoimmune diseases.^34–36 Data from animal models indicate that cigarette smoking also increases expression of matrix metalloproteinase 9.³⁷ Increases in the ratio of matrix metalloproteinase 9 to the levels of its endogenous inhibitors are associated with gadolinium CE lesion activity on MRI of patients with MS.³⁸ Cigarette smoking also has antiestrogenic effects through the formation of inactive 2-hydroxy catechol estrogens³⁹ and affects the hormonal balance in women, which may in turn affect the Th1–Th2 balance.⁴⁰

Our findings provide evidence for a link between smoking and brain injury and could be important for providing the support for antismoking education in schools and more targeted smoking cessation programs to patients with MS. The potential mechanisms are still unclear, but the growing body of epidemiologic evidence for the association of cigarette smoking and MS warrants further investigation with well-designed prospective studies with MRI measures.

AUTHOR CONTRIBUTIONS

Dr. Robert Zivadinov contributed to study design, MRI and clinical data acquisition and analysis, and manuscript preparation. Dr. Bianca Weinstock-Guttman contributed to the clinical and neurologic aspects, patient consent and recruitment, and manuscript preparation. Murali Ramanathan contributed to data analysis, result interpretation, and manuscript preparation. All other authors contributed to MRI image analysis.

ACKNOWLEDGMENT

The authors thank Wee V. Yong, PhD, Hotchkiss Brain Institute and Department of Clinical Neurosciences, University of Calgary, Alberta, Canada, for useful input for this manuscript.

DISCLOSURE

Dr. Zivadinov has received speaker honoraria and consultant fees from Teva Neurosciences, Biogen Idec, Aspreva, Pfizer, and EMD Serono; and has received research support from the National Multiple Sclerosis Society, the National Science Foundation, Biogen Idec, Teva Neuroscience, Genzyme, Aspreva, and Jog for the Jake Foundation. Dr. Weinstock-Guttman has received speaker honoraria and consultant fees from Biogen Idec, TEVA Neurosciences, EMD Serono, and Novartis; and has received research support from Biogen Idec, EMD Serono, TEVA, Cyberonics, and the National Multiple Sclerosis Society. Dr. Hashmi, Dr. Abdelrahman, Dr. Stosic, Mr. Dwyer, Ms. Hussein, and Ms. Durfee report no disclosures. Dr. Ramanathan serves as an editor for the American Association of Pharmaceutical Scientists Journal; receives royalties for publishing The Pharmacy Calculations Workbook [Pinnacle, Summit and Zenith, 2008]; and receives research support from EMD Serono, Novartis, Pfizer, the National Multiple Sclerosis Society, and the National Science Foundation.