Cerebral hypoperfusion: a new pathophysiologic concept in multiple sclerosis?

Abstract

The exact pathogenesis of multiple sclerosis (MS) is incompletely understood. Although auto-immune responses have an important role in the development of hallmark focal demyelinating lesions, the underlying mechanism of axonal degeneration, the other key player in MS pathology and main determinant of long-term disability, remains unclear and corresponds poorly with inflammatory disease activity. Perfusion-weighted imaging studies have demonstrated that there is a widespread cerebral hypoperfusion in patients with MS, which is present from the early beginning to more advanced disease stages. This reduced cerebral blood flow (CBF) does not seems to be secondary to loss of axonal integrity with decreased metabolic demands but appears to be mediated by elevated levels of the potent vasospastic peptide endothelin-1 in the cerebral circulation. Evidence is evolving that cerebral hypoperfusion in MS is associated with chronic hypoxia, focal lesion formation, diffuse axonal degeneration, cognitive dysfunction, and fatigue. Restoring CBF may therefore emerge as a new therapeutic target in MS.

Introduction

Multiple sclerosis (MS) is a leading cause of neurologic disability in young to middle-aged adults and therefore has an important socio-economic impact.¹ The majority of patients start with a relapsing–remitting pattern in which exacerbations of neurologic dysfunction, which generally evolve over days to weeks, are followed by periods of at least partial remission. Many of them eventually convert to a secondary progressive phase characterized by a continuous downhill course that may still be accompanied by some superimposed relapses. There is also a primary progressive form of MS characterized by a steady progression of neurologic disability from disease onset with no or occasional relapses.²

As T-cell mediated inflammatory responses against myelin of the central nervous system are involved in the formation of typical focal demyelinating lesions, which constitute the pathologic substrate for relapses, MS has traditionally been considered as an auto-immune disorder.^{3, 4} However, the self-antigen still remains to be determined and some observations suggest that MS pathology may not exclusively originate from primary immune dysregulation. First, a subset of focal MS lesions can develop without a preceding inflammatory response and appears to be caused by a primary oligodendrogliopathy with extensive apoptosis and microglial activation.^{5, 6} Second, the underlying mechanism of axonal degeneration, the key player in progressive disease and main determinant of long-term disability in MS, remains unclear and seems to run independently from inflammatory activity.⁷ Third, a definite cure for MS is still lacking despite the fact that few neurologic disorders have seen so many new management options emerge over the past 20 years as MS, moving the disease from being untreatable to one with a wide range of available oral and injectable therapies. Current disease-modifying treatments include interferons, glatiramer acetate, teriflunomide, fingolimod, dimethylfumarate, natalizumab, alemtuzumab, and mitoxantrone. All these drugs interfere with the immune system and reduce relapses, but none have proven to be beneficial in progressive MS. Neurodegeneration even continues despite interventions that profoundly suppress the immune system, including humanized antileucocyte monoclonal antibody treatment (alemtuzumab) and autologous hematopoetic stem cell transplantation.^{8, 9}

Recent perfusion-weighted imaging studies have demonstrated that there is a globally decreased cerebral perfusion in subjects with MS, which is independent of the disease course (relapsing–remitting or progressive onset) and does not seem to be driven by axonal degeneration with reduced metabolic demands.¹⁰ This review focuses on the role of cerebral hypoperfusion as a potential new pathophysiologic and therapeutic concept in MS.

Reduced cerebral blood flow (CBF) in MS

Cerebral perfusion is usually quantified as CBF, which represents the blood volume that passes through a given volume of brain parenchyma per time unit.¹¹ Reduced cerebral perfusion in both the gray and white matter of patients with MS has already been demonstrated about 30 years ago by single-photon emission computed tomography and positron emission tomography.^{12, 13, 14, 15} However, these studies suffered from low spatial resolution and received little attention at the time.

The topic regained interest during the past decade with the development of more accurate imaging and processing techniques, allowing better visualization and differentiation between white matter plaques, normal-appearing white matter (NAWM) and gray matter. By using dynamic susceptibility contrast-enhanced perfusion magnetic resonance imaging (DSC-MRI), investigators at the New York University found that CBF was decreased in the NAWM of subjects with relapsing–remitting MS, primary progressive MS, and clinically isolated syndromes suggestive for MS, as compared with healthy controls.^{16, 17, 18} Similar observations were made in the deep gray matter of clinically isolated syndromes suggestive for MS and relapsing–remitting MS patients.¹⁹ These results suggest that cerebral hypoperfusion, regardless of the clinical subtype, is an early and integral part of MS pathology.

A study using arterial spin labeling, which is a non-invasive perfusion-weighted MRI method, found an increased white matter CBF in patients with relapsing–remitting and secondary progressive MS, as compared with controls.²⁰ This might be explained by the fact that the authors did not distinguish NAWM from focal white matter lesions. Gadolinium-enhancing areas show increased CBF on DSC-MRI,^{21, 22, 23} and total white matter CBF may thus be overestimated when focal lesions were not properly excluded from perfusion measurements. It is worth mentioning that the same arterial spin labeling study reported decreased thalamic perfusion, as compared with controls, in patients with relapsing–remitting, secondary progressive, primary progressive, and so-called benign MS.²⁰

Based on ultrasound and selective venography studies, it has been proposed that a chronically impaired venous drainage from the central nervous system, for which the term chronic cerebrospinal venous insufficiency has been coined, might have a role in the pathogenesis of MS;^{24, 25, 26} a controversial statement that has stirred the MS community but for which direct evidence is lacking.¹⁰ High-field susceptibility weighted MRI findings argue against the presence of venous stasis in MS,²⁷ and structural anomalies of intra- and extracranial venous anatomy were observed with similar frequency in patients with MS and healthy controls.²⁸ In a recent phase-contrast MRI study by El Sankari et al.,²⁹ venous cerebral and cervical outflow was comparable between patients with MS and control subjects, but the authors did report a significantly decreased total cervical arterial blood flow in the patients with MS, further illustrating that reduced blood supply to the brain is the real vascular problem in MS.

Possible causative mechanisms

Astrocytes are the cells that actively regulate cerebral blood supply to match regional neuronal glucose and oxygen needs,³⁰ but the axonal loss that occurs in patients with MS does not seem to be the driving mechanism of cerebral hypoperfusion. Saindane et al.³¹ investigated the relation between perfusion-weighted and diffusion tensor MRI features in the normal-appearing corpus callosum of patients with relapsing–remitting MS. Decreased CBF positively correlated with mean diffusivity but not with fractional anisotropy, a finding consistent with primary ischemia, rather than with hypoperfusion secondary to axonal degeneration.³¹ In addition, reduced cortical and deep gray matter CBF is already present in patients with early relapsing–remitting MS in the absence of corresponding volume loss.³² Magnetic resonance spectroscopy studies have demonstrated decreased levels of N-acetylaspartate (NAA), a marker for reduced axonal metabolism,^{33, 34} in the normal-appearing centrum semiovale of patients with MS.^{35, 36} Cerebral perfusion positively correlated with NAA levels in healthy controls but this was lost in patients with MS, who had a perfusion reduction greater than would be expected from decreased axonal metabolism or axonal loss alone.³⁷ Moreover, in patients with progressive MS there is an increased excitability of primary motor cortex neurons and therefore a potential increase of their metabolic demand.^{38, 39, 40}

Another hypothesis has associated reduced blood flow with obliterating perivascular MS lesions, but this seems unlikely because one would then expect a more patchy pattern of focal CBF decrease, as seen in central nervous system vasculitis, which is not the case in MS. Microvessel thrombosis and other structural abnormalities have only very exceptionally been observed within MS plaques.⁴¹ The increased CBF in active inflammatory lesions also argue against this theory.

Hypercapnic perfusion MRI in patients with MS has demonstrated an impaired dilatory capacity of cerebral arterioles in response to vasomotor stimulation.⁴² Enhanced blood levels of the potent vasoconstrictive peptide endothelin-1 (ET-1) were found in peripheral venous blood and cerebrospinal fluid of patients with MS.^{43, 44} This has been associated with reduced extra-ocular blood flow velocities.⁴⁵ In a recent study, CBF values, measured with arterial spin labeling, were globally (i.e., in NAWM of centrum semiovale, frontoparietal cerebral cortex, thalamus, and cerebellar hemispheres) reduced by approximately 20% in patients with MS, as compared with control subjects. After the administration of the ET-1 receptor antagonist bosentan, CBF in the MS patients significantly increased in all brain regions to reach values obtained in the controls.⁴⁶ Elevated plasma ET-1 levels in both peripheral and internal jugular venous blood and elevated jugular/peripheral vein ratio in MS patients compared with controls suggest that, in MS, an excess of ET-1 is released from the brain into the cerebral circulation.⁴⁶ ET-1 is widely distributed in the human body and mostly produced by endothelial cells.⁴⁷ Under normal circumstances, ET-1 can be found in some type of neurons but not in glial cells. Very little is known about ET-1 in astrocytes, although a few studies have found increased levels in reactive astrocytes in various brain pathologies, such as acute ischemic stroke, Alzheimer's disease, viral infections, and traumatic injury.^{48, 49} ET-1 immunohistochemistry on postmortem white matter brain samples showed that reactive astrocytes in MS plaques stained positive for ET-1, whereas this was not the case for NAWM astrocytes in MS patients and healthy controls.⁴⁶ The above data suggest that reduced CBF in MS is mediated by elevated levels of ET-1, which are likely released in the cerebral circulation from reactive astrocytes in MS plaques to induce arteriolar vasoconstriction.

Role of reduced CBF in MS pathology?

Mitochondrial energetic failure and oxidative stress become increasingly recognized as factors associated with axonal degeneration in MS.^{35, 50} Animal models demonstrate that chronic brain hypoperfusion induces mitochondrial dysfunction and production of free radicals, resulting in neuronal damage.⁵¹ In human MS cortex, evidence exists of significant reductions in gene products specific for the mitochondrial electron transport chain and that cortical mitochondria have a diminished capacity to exchange electrons in respiratory chain complex I and III.⁴⁰ The mitochondrial electron transport chain ionic cascade has been unequivocally documented as the cause of myelinated axonal degeneration in experimental models of central nervous system white matter hypoxia and ischemia.^{40, 52} Cerebral white matter axons and myelin appear to be particularly susceptible to chronic hypoxia.⁵³ Expression of hypoxia-inducible factor-1α and its downstream genes, a salvage pathway that becomes activated in response to tissue hypoxia,⁵⁴ is enhanced in the NAWM of patients with MS.⁵⁵ A similar upregulation has been found in the white matter of subjects with ischemic leukoencephalopathy,⁵⁶ a prototypical disease of reduced white matter CBF.⁵⁷

Elevated levels of hypoxia-inducible factor-1α, as well as p53, another hypoxic stress protein,⁵⁸ have also been observed in MS plaques.^{41, 59} Luchinetti et al.⁶⁰ have distinguished four histologically different demyelination patterns. Type III lesions, which are characterized by preferential loss of myelin-associated glycoprotein and oligodendocyte apoptosis, and demyelination with little or no inflammation, show remarkable similarity with classic vascular white matter injury.⁶¹ It has been suggested that this type III pattern may be characteristic for early lesions, whereas immune activation appears in later stages.⁶² This theory is not supported by DSC-MRI findings of elevated CBF in acute gadolinium-enhancing lesions, compared with surrounding white matter tissue, attributed to inflammation-mediated vasodilation.^{21, 22, 23} However, in a subset of lesion that developed ring enhancement increased perfusion was only seen in the ring tissue, whereas inside the ring there was a CBF decrease suggestive for central ischemia.²³ Moreover, some MS patients develop new plaques with reduced MRI mean diffusivity, similar to what is observed in acute ischemic stroke.⁶³

Interestingly, it has been found that hyperintense T2-weighted MRI lesions in patients with MS are situated preferentially in lower perfused white matter areas,⁶⁴ and lesion volume negatively correlated with regional CBF.⁶⁵ A recent study by Narayana et al.⁶⁶ showed that T1-hypointense lesions occurred almost exclusively in brain regions with lower CBF, whereas gadolinium enhancement was distributed more homogeneously according to perfusion. Persistent T1-hypointense areas or black holes are generally considered as irreversible tissue damage and indicators of poor clinical prognosis.⁶⁶ These results suggest that reduced CBF may contribute to focal lesion formation.

From a clinical point of view, reduced white and gray matter CBF in patients with MS has thus far been associated with cognitive manifestations,^{67, 68, 69, 70} which are remarkably similar to those described in ischemic leukoencephalopathy.⁷¹ One study found that reduced deep gray matter perfusion in MS negatively correlated with fatigue,⁷² which is a very common but poorly understood cause of impaired activities of daily living in patients with MS. Fatigue in MS is often referred to as central fatigue because there is now strong evidence to suggest that fatigue results from reduced voluntary activation of muscles by means of central mechanisms.⁷³ Global disability or the rate of disease progression, as measured by Expanded Disease Severity Status (EDSS) scores or Multiple Sclerosis Severity Scores (MSSS), respectively, do not seem to be influenced by impaired CBF,³⁷ but this may be because of the fact that cognitive dysfunction and/or central fatigue cannot be properly evaluated by EDSS or MSSS scales. An additional reason for the absence of correlation may be the influence of spinal lesions on EDSS and MSSS.

Conclusions and future directions

Cerebral hypoperfusion is an early an integral feature of MS pathology that may be more relevant than hitherto thought. Reduced CBF in patients with MS is not the result of loss of axonal integrity but appears to be mediated by elevated levels of ET-1, likely released in the cerebral circulation by reactive astrocytes, and, moreover, can be reversed by the non-selective ET-1 receptor antagonist bosentan. Evidence is evolving that cerebral hypoperfusion in MS might have a role in the formation of a subset of focal lesions, axonal degeneration, cognitive dysfunction, and fatigue. However, definite proof will require well-designed clinical studies assessing the effect of long-term restoration of cerebral perfusion on these clinical parameters. Another interesting but unresolved question concerns the underlying mechanism of ET-1 upregulation in reactive MS astrocytes. Interestingly, ET-1 expression by reactive astrocytes has also been described in Alzheimer's disease and chronic ischemic leukoencephalopathy, which are other conditions associated with reactive astrocytes expressing ET-1 immunoreactivity and a global CBF decrease.⁷⁴ ET-1 transcription may be enhanced by cytokines that are elevated in focal MS lesions, including tumor necrosis factor-α and interleukin-1β.^{75, 76, 77} Astrocytes in MS are deficient in β₂-adrenergic receptors, which regulate intracellular levels of cAMP.⁷⁸ A possible association between β₂-adrenergic receptor deficiency in MS astrocytes remains highly hypothetical, but cAMP was found to inhibit basal ET-1 production in other cell types.⁷⁹ In vitro experiments are warranted to determine the effects of these cytokines and β₂-adrenergic receptor modulation on astrocyte ET-1 production.

The authors declare no conflict of interest.