Stroke prognosis in diabetes mellitus: new insights but questions remain

Diabetes mellitus accelerates the clinical course of atherosclerosis and increases both cardiovascular morbidity and mortality [1]. Besides hypertension and dyslipidemia, diabetes is a well-known independent risk factor for coronary heart disease and ischemic stroke, particularly atherothrombotic infarction and lacunes [2]. In a classical autopsy study of acute cerebrovascular accidents in diabetes, cerebral small-vessel disease was 2.5-times more frequent in diabetes [3]. In clinical series of lacunar infarctions, the incidence of diabetes varied between 11 and 29% [4–7]. Diabetes is not only an independent risk factor for lacunar infarction, especially in patients with multiple lacunar infarcts [7], but the presence of diabetes is also associated with worse functional recovery in these patients [8]. In a recent study, diabetes and hypertension were independent factors related to the recurrence of lacunar stroke in patients who had suffered from a previous lacunar stroke [9]. Moreover, cognitive impairment was observed in 16% of patients with a first lacunar stroke recurrence and in 40% of those with multiple recurrent lacunes [9].

Although the deleterious effect of diabetes on cardiovascular morbimortality has been investigated extensively, a limited number of prospective studies have examined in-hospital mortality in patients with ischemic stroke and diabetes. In the study by Jorgensen et al., in-hospital mortality in stroke patients with diabetes was 24% compared with 17% in those without diabetes; diabetes increased the relative death risk by 1.8 [10]. The reason for the increased early death and worse prognosis associated with cerebral infarction in diabetic patients is not well understood. Diabetic stroke patients had atrial fibrillation more often than nondiabetic subjects with stroke [11]. Congestive heart failure and atrial fibrillation constitutes a major aggravating factor in this population and is likely to further increase the risk of stroke, cardiac events and sudden death substantially, also suggesting that diabetic-stroke patients may have larger cerebral infarcts [11,12]. In one study of cerebral infarction in diabetes, independent clinical factors related to in-hospital mortality in diabetic patients with ischemic stroke were age, atrial fibrillation, congestive heart failure, chronic nephropathy and altered consciousness [12]. Conversely, women with diabetes constitute a subgroup of stroke patients with a high risk for fatal outcome. The increased stroke mortality in postmenopausal diabetic female patients would be mainly caused by the higher occurrence of hypertension and cardiac disorders, in particular, congestive heart failure and atrial fibrillation and advanced age, both of which are indicators of in-hospital mortality after stroke in general series of cerebrovascular disease patients [13,14].

Hyperglycemia also has a deleterious effect on intracerebral hemorrhage. In an experimental rat model of collagenase-induced, intracerebral hemorrhage, hyperglycemia caused more profound brain edema and perihematomal cell death in hyperglycemic rats than in the normoglycemic group [15]. In a clinical study of patients with acute spontaneous intracerebral hemorrhage, diabetes was the vascular risk factor associated with higher in-hospital mortality [16]. Patients with diabetes presented hematoma of multiple topography and previous cerebral infarction, factors that can be associated with higher in-hospital mortality. Intracerebral hemorrhage of large size in diabetic patients may be related to the specific small-vessel disease induced by diabetes and characterized by lipohyalinosis, fibrinoid necrosis, microatheroma and microaneurysms [6].

Elevated blood glucose is common in the early phase of stroke. The prevalence of hyperglycemia, defined as a blood glucose level of more than 6.0 mmol/l (108 mg/dl), has been observed in two thirds of all ischemic stroke subtypes on admission. Although up to a third of acute stroke patients have either previously diagnosed or newly diagnosed diabetes, probably a major proportion of patients have stress hyperglycemia, partly mediated by the release of cortisol and norepinephrine [17]. Acutely elevated hyperglycemia is associated with poor outcomes in stroke patients of both genders, with and without history of diabetes, in relation to larger infarct volume and cortical involvement. In a recent multicenter, prospective study of patients with acute ischemic stroke, high capillary glucose levels of greater than 155 mg/dl (8.5 mmol/l) at any time within the first 48 h was associated with a higher risk of death and poor outcome, independent of age, stroke severity, diabetes or infarct volume [18].

Mechanisms involved in hyperglycemia-mediated cerebral damage include the liberation of lactic acid and free radicals, dysfunction of mitochondrial activity, increased edema formation in the penumbral tissue, microvascular disruption of the blood–brain barrier with erythrocyte extravasation and subsequent hemorrhagic transformation [12,14,17]. Magnetic resonance spectroscopy studies have also shown that acute hyperglycemia increases brain lactate production and facilitates the conversion of hypoperfused at-risk tissue into infarction [19]. In addition, diffusion tensor imaging is a sensitive method to identify the number and volume of clinically silent lesions.

The use of intravenous saline and avoidance of glucose solutions in the first 24 h after stroke appears to reduce glucose levels. Hypoglycemia (<50 mg/dl [< 2.8 mM]) may mimic an acute ischemic stroke and should be treated by intravenous dextrose bolus or infusion of 10–20% glucose. In cases of serum glucose levels greater than 180 mg/dl (> 10 mM), the administration of insulin is recommended [20]. Conversely, blood glucose should be checked regularly. Diabetes should be managed with lifestyle modification and individualized pharmacological therapy. Patients with Type 2 diabetes who do not need insulin may benefit from treatment with pioglitazone for preventing the occurrence of ischemic stroke [21].

An interesting aspect refers to the relevance of hyperglycemia in stroke patients who are candidates for intravenous thrombolytic therapy. Admission hyperglycemia emerged as a robust predictor of:

• • Increased risk of hemorrhagic transformation after the administration of thrombolytic agents, because early hyperglycemia is associated with hemorrhagic transformation in stroke [17]

• • Thrombolysis resistance regardless of occlusion location

Acute hyperglycemia may hamper the recanalization process through mechanisms including the inhibition of plasma fibrinolysis, increasing plasminogen activator inhibitor type 1, and decreasing tissue plasminogen activator (tPA) activity. A recent study of 221 stroke patients treated with intravenous tPA provided evidence of the role of high glucose levels in delaying reperfusion of the brain by impairing tPA-induced recanalization [22]. Hyperglycemia greater than 140 mg/dl and the presence of tandem internal carotid artery and/or middle cerebral artery occlusion emerged as independent predictors of absence of recanalization. In distal middle cerebral artery occlusions, hyperglycemia was also the only predictor of absence of recanalization.

Conversely, several lines of investigation suggest a link among diabetes, cognitive impairment and dementia. Clinical studies have shown impaired neuropsychological functioning in patients with diabetes. Compared with community-dwelling, normoglycemic individuals, those with diabetes have a higher prevalence of global cognitive impairment and a higher incidence of cognitive decline. Population-based studies have also shown that diabetes is a risk factor for Alzheimer’s disease [23]. An increasing body of data supports the hypothesis that diabetic pathologies lead to both Alzheimer-type neurodegeneration and vascular brain pathology, and it is this mix of these diseases that forms the anatomical basis for clinical and subclinical cognitive impairment in diabetes. Diabetes is associated with both vascular disease (i.e., infarcts) and neurodegenerative changes (i.e., hippocampal atrophy), which are frequently seen in Alzheimer’s disease. These data suggest that, compared with nondiabetic patients, those with diabetes have brain structural changes that reflect neuronal degeneration as well as vascular damage [23].

It is well known that diabetes is an atherogenic risk factor that can disrupt vascular reactivity via the formation of advanced glycation end products (AGEs), resulting in subtle perfusion abnormalities. The increased risk of Alzheimer’s disease may also be mediated by the exacerbation of β-amyloid neurotoxicity by AGEs, which have been identified in the matrix of neurofibrillary tangles and amyloid plaques in brains from subjects with Alzheimer’s disease, or alternatively, associated with insulin functions [24]. Insulin resistance appeared to be a midlife risk factor for cognitive decline and dementia [25]. Longitudinal studies provide evidence that diabetes mellitus is a risk factor for cognitive decline and dementia (Alzheimer’s disease and vascular-type dementia), but the pathophysiological mechanisms underlying this association are still unknown.

Three main pathological mechanisms have been proposed to explain the link between diabetes and dementia [26,27]. First, diabetes may lead to dementia through ischemic cerebral small-vessel disease, especially in elderly people [28]. Second, hyperglycemia may have a direct toxic effect on neurons by causing oxidative stress and the accumulation of AGEs, a process that can directly affect brain tissue and also lead to microvascular changes [29]. Third, insulin and insulin-degrading enzymes could play an important role in amyloid metabolism [27]. As for diabetes, excess adiposity could increase the risk of dementia by a vascular pathway [30]. Furthermore, adipose tissue secretes adipocytokines, such as leptin, which may be involved in neurodegenerative pathways [31]. The metabolic syndrome is a cluster of cardiovascular risk factors, including obesity and insulin resistance, and it has been proven to be associated with an increased risk of developing cardiovascular disease. It is also an independent risk factor for silent brain infarctions, identified using MRI, in middle-aged, healthy people [32].

Cerebrovascular lesions, especially small-vessel disease, due to diabetes may magnify the effect of mild Alzheimer’s disease pathology and promote the progression of cognitive decline, and also be a precursor of neuronal damage and dementia. Therefore, we believe that, instead of labeling all patients as either neurodegenerative or cerebrovascular, both disorders could be viewed as a continuum, with purely neurodegenerative disease at one end and purely cerebrovascular disease at the other end of the spectrum [34]. Recent studies have suggested a vascular origin of Alzheimer’s disease, based on the decreased blood flow that causes the cerebral hypoxia and posterior brain degeneration [34,35]. We predict that managing Alzheimer’s disease as a vascular disease will open a new window for an improved understanding of the etiopathogenesis and lead to the discovery of new and more effective treatments.

Recently, lacunar disease has been considered a focal manifestation of a pathologically diffuse and progressive vascular disease of the cerebral arterioles of small calibre, with the possibility of causing cognitive impairment and dementia when lacunar disease is sufficiently extensive [8]. Therefore, cerebral small-vessel endothelial dysfunction, facilitated by hypertension and diabetes, would cause: blood–brain barrier leakage, with extravasation of substances that do not normally enter the brain interstitial space (e.g., plasmin and other proteases); direct arteriolar wall, perivascular tissue and, neuronal and glial damage [36]; and narrowing of lumen leading to reduced blood flow and ischemia. This asymptomatic and gradual process may explain the asymptomatic progression of lacunar disease and the development of silent lacunes, leukoaraiosis, progressive cognitive impairment and dementia [6]. In this scenario, retinal microvascular changes observed in the study of Lindley et al. represent a directly visible microvascular phenotype of endothelial dysfunction and would support the validity of the aforementioned alternative physiopathological hypothesis [37,38].

It should be emphasized that results obtained from experimental studies have shown that hyperglycemia exacerbates the ischemic lesions by increasing acidosis-related damage and is associated with an increase of the cerebral edema and size of the infarct [1,39]. Cerebral edema is a complex pathophysiological process that causes brain swelling, complicates ischemic stroke, worsens neurological function and can lead to brain herniation and death [40]. It has been shown that nonhypoglycemogenic low doses of glibenclamide have a strong beneficial effect on lesion volume and also have a highly favorable therapeutic window in several models of ischemic stroke [41]. Recent experimental studies have shown that sulfonylureas may have a beneficial effect on cerebral edema [42]. Sulfonylurea derivatives constitute the pharmacological class of oral hypoglycemic agents most frequently used in the treatment of Type 2 diabetes. Sulfonylurea derivatives act by depolarizing pancreatic β cells by inhibiting ATP-dependent potassium channels (K_ATP) [43,44]. Simard et al. recently identified a nonselective cation channel (NC): the (Calcium-ATP) channel, in ischemic astrocytes which is regulated by sulfonylurea receptor 1 (SUR1), opened by depletion of ATP, and, when opened, causes cytotoxic edema, oncotic cell death and cerebral edema [42]. Similar to the K_ATP channel in pancreatic β cells, the NC (Ca-ATP) channel is regulated by SUR1 and is blocked by sulfonylureas. Simard et al. found that this channel is upregulated in rodent models of ischemic stroke, and blockage of SUR1 with constant infusion of low-dose glibenclamide only caused a slight reduction of serum glucose but was highly effective in reducing cerebral edema, infarct volume and mortality by 59% in rodent models of stroke [42]. Accordingly, the NC (Ca-ATP) channel is crucially involved in development of cerebral edema, and targeting SUR1 may provide a new therapeutic approach to stroke.

These novel, promising results obtained in experimental studies have been replicated in humans. Kunte et al. showed for the first time that treatment with sulfonylureas before cerebral ischemia and maintained during the acute phase of infarction (similarly to therapy in experimental studies) had a beneficial effect on the short-term prognosis of patients with Type 2 diabetes and cerebral infarction [45]. The authors studied a cohort of 33 diabetic patients with cerebral ischemia who were treated with glibenclamide, glimepiride or glibornuride on admission through discharge (treatment group) and 28 diabetic patients with cerebral ischemia not taking a sulfonylurea drug (control group). A decrease in NIH Stroke Scale of four or more points from admission to discharge, or a discharge NIH Stroke Scale score of 0 was reached by 36% of patients in the treatment group and 7.1% in the control group (p = 0.007). At the time of discharge, modified Rankin scale score of up to two was obtained by 81.8% in the treatment group and 57.1% in the control group (p = 0.035). Improvement was independent of gender and previous transient ischemic attack and blood glucose levels, and it is remarkable that it was only observed in the subset of patients with cerebral infarction of nonlacunar type. Similarly, previous studies assessing the role of statins or HMG-CoA reductase inhibitors have shown that patients who are treated with a statin before or early after an ischemic stroke have a more favorable outcome than those who are not [46,47]. Previous transient ischemic attack also showed a potentially beneficial effect, probably by a mechanism of ischemic tolerance [48]. In experimental studies, Glazier et al. showed that induction of ischemic tolerance following brief focal ischemia in rat brains was produced in the middle cerebral artery territory of the neocortex ipsilateral to the ischemic preconditioning [49]. By contrast, the extent of neuronal necrosis in subcortical structures was similar in both hemispheres. The lack of efficacy of sulfonylurea derivatives in lacunar infarction may be explained by the small size of lacunar infarcts in which the maximum diameter of the lesion is less than 15 mm and the well-known, short-term good prognosis of this stroke subtype [8], so that it is difficult to quantify differences in neurological recovery in this subgroup of patients who spontaneously present a good outcome. Another phenomenon that might be related to the better prognosis of lacunar ischemic stroke, such as the preparatory recruitment of collateral pathways, is not possible in lacunar infarction, because this type of cerebral ischemia is characterized by occlusion of the terminal, perforating arterioles with no possibility of prior mobilization of collateral pathways [8,46].

Effective therapies for acute stroke are scarce, and numerous neuroprotective strategies have failed in human trials [50]. Glibenclamide seems to be useful because it is highly effective at blocking cellular edema and ionic edema that, together, are the main determinants of brain swelling.

Conclusion

Although experimental data and findings in recent clinical studies have provided new insights into the impact of diabetes on the prognosis of stroke, unanswered questions still remain. Areas for future research include the impact of diabetes as a prognostic factor of poor outcome in primary intracerebral hemorrhage, the relationship between diabetes and cerebral neurodegenerative processes, the potential neuroprotective role of sulfonylurea derivatives, and the beneficial short- and long-term effect of these agents in diabetic stroke patients.

Financial & competing interests disclosure

Adrià Arboix belongs to CIBER de Enfermedades Respiratorias (CB06/06), Institute Carlos III, Madrid, Spain. The study was supported, in part, by a grant from FIS PI081514, Madrid, Spain. The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Writing assistance was utilized in the production of this manuscript. The author would like to thank Marta Pulido, MD, for editing the manuscript and for editorial assistance.