General Management of Elevated Blood Pressure, Blood Glucose, and Body Temperature
Monitoring the blood pressure (BP), glucose levels, and temperature in acute stroke patients is an often neglected matter, although it may have an important impact upon the patient’s outcome. In the Tel Aviv stroke register, recorded between the years 2001 and 2003, 32% of acute stroke patients in the emergency room had glucose levels higher than 150 mg/dl, systolic BP higher than 140 mmHg was found in 77% of the patients, and 17% of patients had temperatures above 37°C on admission. These numbers are representative of other centers as well. This chapter will summarize the current knowledge regarding the management of the above.
Hypertensive Blood Pressure Values in Acute Ischemic Stroke
Several observations have demonstrated spontaneous elevation of BP in the first 24–48 hours after stroke onset with a significant spontaneous decline after a few days [1–3]. Several mechanisms may be responsible for the increased BP, including stress, pain, urinary retention, and Cushing effect due to increased intracranial pressure and the activation of the sympathetic, renin–angiotensin, and adrenocorticotropic (ACTH)–cortisol pathways. Carlberg et al. did not find correlation between the admission hypertension and time from stroke onset; they concluded that mental stress was responsible for the increase of BP level after stroke . Also, positive association between salivary cortisol and 24-hour BP supports the theory of the stress response as a determinant of BP levels in acute stroke . Despite the increased prevalence of hypertension following stroke, optimal management of BP in acute stroke has not been yet established. It is important to emphasize that it is not yet known whether BP changes after stroke are physiological and beneficial reaction to stroke or they simply reflect either mental stress reaction to stroke or hospital admission . Several arguments speak for lowering the elevated BP: risks of hemorrhagic transformation, cerebral edema, recurrence of stroke, and hypertensive encephalopathy. On the other hand, it may be important to maintain the hypertensive state due to the damaged autoregulation in the ischemic brain and the risk of cerebral hypoperfusion exacerbated by the lowered systemic BP.
Blood Pressure and Outcome
Analysis of 17 398 patients in the International Stroke Trial  demonstrated a U-shaped relationship between baseline systolic BP and both early death and late death or dependency. Both high BP and low BP were independent prognostic factors for poor outcome. Early death increased by 17.9% for every 10 mmHg below 150 mmHg (p <0.0001) and by 3.8% for every 10 mmHg above 150 mmHg (p <0.016). A prospective study among 1 121 patients admitted within 24 hours from stroke onset and followed up for 12 months demonstrated similar findings of the “U-shape” phenomenon . Elevated pulse pressure (difference between the systolic and diastolic BP) during the acute phase of ischemic stroke was also found to be an independent predictor of poor early outcome at hospital discharge and 30-day mortality .
It should be taken into consideration that prolongation of the elevated BP may be caused by more severe stroke as compensation for the persistent vessel occlusion.
On the other hand, the GAIN study , done among 1 455 patients with ischemic stroke, demonstrated that baseline mean arterial pressure was not associated with poor outcome. However, variables describing the course of BP over the first days have a marked and independent relationship with 1- and 3-month outcomes.
It is important to mention Christensen’s study, which showed that decrease in BP value during the first 4 hours after admission was associated with milder stroke with favorable outcome, while a maintained high BP was associated with severe stroke and poor outcome .
Many authors suggest that high admission BP in patients with moderate stroke may be an impact of hospitalization. Jansen et al. and Semplicini et al. showed a return of BP to a lower level than that of pre-stroke 3 days after admission, followed by a BP that is equal to the pre-stroke values 3 months after stroke onset. Also, two peaks of BP elevation were noticed – first in the emergency room and the second after admission to the neurological department, which is probably the result of the two admissions [10, 11].
In a Cochrane systematic review of 32 studies involving 10 892 patients after ischemic and hemorrhagic stroke , death was found to be significantly associated with elevated mean arterial BP (odds ratio [OR] 1.61; 95% confidence interval [CI] 1.12–2.31) and high diastolic BP (OR 1.71; 95% CI 1.33–2.48).
A U-shaped relationship between baseline systolic BP and both early and late death or dependency after ischemic stroke has been demonstrated in clinical trials.
Blood Pressure and Outcome in Thrombolyzed Patients
Several observations, including the National Institute of Neurological Disorders and Stroke (NINDS) rtPA trial [13, 14], found an association between high BP on admission, and its prolongation, with poor outcome and mortality. Although in one study no such association was found in alert patients, stroke patients with impaired consciousness showed higher mortality rates with increasing BP . The Safe Implementation of Thrombolysis in Stroke (SITS) thrombolysis register prospectively recorded 11 080 stroke patients treated with intravenous thrombolysis. BP values were recorded at baseline, 2 hours, and 24 hours after thrombolysis . High systolic BP was associated with poor outcome. Withholding anti-hypertensive therapy up to 7 days in patients with a history of hypertension was associated with worse outcome, whereas initiation of anti-hypertensive therapy in newly recognized moderate hypertension was associated with a favorable outcome.
The association between elevated BP and recanalization was evaluated in 149 patients after intra-arterial thrombolysis using angiography . The study demonstrated that the course of elevated systolic BP, but not diastolic BP, after acute ischemic stroke was inversely associated with the degree of vessel recanalization. When recanalization failed, systolic BP remained elevated longer than when it succeeded.
In patients treated with intravenous tissue plasminogen activator (IV-tPA), current expert consensus guidelines recommend BP<180/105 mmHg . However, this is based on extrapolating the findings from myocardial infarction thrombolysis trials, largely applicable to BP management in stroke patients with IV-tPA treatment . The data from Safe Implementation of Thrombolysis in Stroke (SITS) study suggest maintaining systolic BP in the range 141–150 mmHg after IV-tPA therapy . The Enhanced Control of Hypertension and Thrombolysis Stroke (ENCHANTED) study started in 2012, and compares systolic BP 130–140 mmHg and systolic BP below 180 mmHg in stroke patients treated with IV-tPA. This study is ongoing and may help resolve the doubt [19, 21]. Despite the mentioned results, the optimal BP management after reperfusion therapy still remains unclear.
Controlling Blood Pressure in the Acute Stroke Phase
The theory that elevated systemic BP may compensate for the decreased cerebral blood flow in the ischemic region led to attempts to elevate BP as a treatment for acute ischemic stroke. The hemodynamic and metabolic impact of pharmacologically increased systemic BP on the ischemic core and penumbra was evaluated in rats. The mild induced hypertension was found to increase collateral flow and oxygenation and to improve cerebral metabolic rate of oxygen in the core and penumbra . Several small studies in humans have addressed this question by investigating the responses to vasopressors, including phenylephrine and norepinephrine, in patients with acute stroke [23–25]. Despite a documented improvement in cerebral blood flow , the concept was abandoned because of the increased risk of hemorrhage and brain edema. In a systematic review of 12 relevant publications including 319 subjects, the small size of the trials and the inconclusive results limit conclusion as to the effects on outcomes – both benefits and harms. A randomized controlled trial is needed to determine the role of vasopressors in acute ischemic stroke . However, in stroke patients with early hypotension due to pre-existing heart disease (poor left ventricular function and low cardiac output, e.g. congestive heart failure, coronary heart disease, arrhythmia) or due to dehydration, aggressive anti-hypertensive therapy, or septic condition (e.g. hidden infection), which can cause infarct progression and neurological deterioration, by decrease of collateral blood flow, increasing systemic BP could save the ischemic penumbra by raising intraluminal hydrostatic pressure, opening collateral channels, and improving perfusion to penumbral tissues [28–30]. However, the question: Should BP be elevated to improve cerebral perfusion in patients with ischemic stroke? The answer is not simple. A few small studies showed improvement of neurological status after induced hypertensive therapy. Also, there are ongoing trials to assess the usefulness of hypertensive therapy in patients with a diffusion–perfusion MRI mismatch. In the meantime, it is rational to try volume expanders and/or vasopressors in patients with hypotensive stroke or in patients who have had a neurological deterioration associated with a BP drop .
Elevated systemic blood pressure may compensate for the decrease of cerebral blood flow in the ischemic region, but raises the risks of hemorrhagic transformation, cerebral edema, recurrence of stroke, and hypertensive encephalopathy.
According to a systematic review of the literature , no conclusive evidence to support the lowering of BP in the acute phase of ischemic stroke was found and more research is needed to identify the effective strategies for BP management in that phase . Despite the controversy over the management of BP in the acute phase, the benefit of BP reduction as a secondary prevention of stroke is well established and has been demonstrated in many studies. However, in most of these studies anti-hypertensive agents were administrated several weeks after stroke onset. Only a few trials were performed in the acute stage. The ACCESS trial  was a prospective, double-blind, placebo-controlled, randomized study evaluating the angiotensin-receptor blocker candesartan versus placebo for 342 hypertensive patients in the first week following stroke. Treatment was started with 4 mg candesartan or placebo on day 1 and dosage was increased to 8 or 16 mg candesartan or placebo on day 2, depending on the BP values. Treatment was aimed at a 10–15% BP reduction within 24 hours. Although no difference was found in stroke outcome at 3 months, a significantly lower recurrent cardiovascular event rate and lower mortality after 1 year were documented in the treatment group. The authors concluded that when there is need for or no contraindication against early anti-hypertensive therapy, candesartan is a safe therapeutic option.
In the UK’s Controlling Hypertension and Hypotension Immediately Post-Stroke (CHHIPS) pilot trial , researchers randomized 179 patients who had suffered ischemic or hemorrhagic strokes within the previous 36 hours and who also had hypertension defined as systolic BP greater than 160 mmHg. Patients received doses of either the anti-hypertensive drugs lisinopril at a dosage of 5 mg or labetalol at a dosage of 50 mg or a placebo for 14 days. Three months after treatment began, the active treatment group had a significantly lower mortality compared to the placebo group. However, the recent SCAST study  showed no indication that BP-lowering treatment with candesartan is beneficial in patients with acute stroke and raised BP (>140 mmHg systolic). If anything, the evidence suggested a harmful effect.
Another important issue for consideration is whether patients who are already on anti-hypertensive treatment should continue or stop their pre-existing drugs. In the Continue or Stop Post-Stroke Anti-hypertensives Collaborative Study (COSSACS) , continuation of anti-hypertensive drugs did not reduce death or dependency at 2 weeks, cardiovascular event rate, or mortality at 6 months. However, due to early termination, the study was underpowered.
It should be considered that stroke patients in these studies have been treated as a homogeneous group, without distinguishing between those who do and do not have hypoperfused brain tissue by perfusion imaging . It stands to reason that the former might benefit from hypertension, whereas the latter would not, and might only suffer the side-effects of elevated BP.
The AHA/ASA guidelines are clear and they recommend restarting anti-hypertensive therapy at 24 hours in previously hypertensive neurologically stable patients unless contraindicated .
Despite the somewhat confusing and unclear data, the current European Stroke Organisation (ESO) 2008 Guidelines  recommend that BP up to 220 mmHg systolic or 120 mmHg diastolic may be tolerated in the acute phase without intervention unless there are cardiac complications. According to the American guidelines , it is generally agreed that patients with markedly elevated BP may have their BP lowered by not more than 15% during the first 24 hours after the onset of stroke. There is an indication to treat BP only if it is above 220 mmHg systolic or if the mean BP is higher than 120 mmHg. No data are available to guide selection of medication for the lowering of BP in the setting of acute ischemic stroke. The recommended medication and doses are based on general consensus.
The Efficacy of Nitric Oxide in Stroke (ENOS) also dealt with similar issues and enrolled 4 011 patients in the period from 2001 to 2013. This study showed that transdermal application of glyceryl trinitrate in patients with acute stroke and high BP, lowered BP and had acceptable safety, but did not improve functional outcome, so there is no evidence to support continuing prestroke anti-hypertensive treatment in patients in the first few days after stroke .
Guidelines recommend blood-pressure-lowering therapy above 220 mmHg systolic blood pressure (ESO 2008 Guidelines and American Guidelines).
It has been well established that elevated glucose levels play a major role in microvascular and macrovascular morbidity and in hematological abnormalities as well. Several processes were found to be associated with these conditions, including impaired vascular tone and flow, disruption to endothelial function, changes at the cellular level, intracellular acidosis, and increased aggregation and coagulability. Some animal studies [40, 41] have demonstrated the relations between acute ischemic stroke and hyperglycemia. In these models the administration of glucose to animals resulted in worsened brain ischemia. Those findings were attributed to the accumulation of lactate, decreased intracellular pH, increase in free radicals and excitatory amino acids, damage to the blood–brain barrier, formation of edema, and elevated risk of hemorrhagic transformation. Pretreatment with insulin was found to limit the ischemia.
As mentioned, 30–40% of acute stroke patients are found to have elevated glucose levels on admission, about half of them have known diabetes, while the others are newly diagnosed or suffer from stress-induced hyperglycemia .
In one systematic study it was shown that glucose pathology is seen in up to 80% of acute patients , many of them showing a high probability of previously unrecognized diabetes. Out of 238 consecutive acute stroke patients, 20.2% had previously known diabetes; 16.4% were classified as having newly diagnosed diabetes, 23.1% as having impaired glucose tolerance (IGT) and 0.8% as having impaired fasting glucose; and only 19.7% showed normal glucose levels.
It is possible that hyperglycemia after stroke is primarily a stress response in relation to stroke severity and size. However, post-stroke hyperglycemia is prevalent across all clinical types and severities of stroke. Even though some authors have suggested that stress hyperglycemia may occur as a result of neuroendocrine dysregulation in response to lesions of insular cortex, this finding has not been consistent in a larger number of studies. It remains unclear whether hyperglycemia after stroke arises as an accompanying phenomenon of stroke in general, as a manifestation of underlying pre-diabetes, or as a consequence of specific anatomic involvement in the brain .
Increased mortality was found in both diabetic and stress-induced hyperglycemia groups, independent of age, stroke type, and stroke size . Stress hyperglycemia was associated with a 3-fold increase in risk of fatal 30-day outcome and 1.4-fold increase in risk of poor functional outcome in non-diabetic patients with acute ischemic stroke. Similar findings were also demonstrated in the NINDS rtPA stroke trial. Hyperglycemia on admission was correlated with decreased neurological improvement and the risk of hemorrhagic transformation in reperfused thrombolyzed patients, but not in non-reperfused rtPA-treated patients . On the other hand, in the NINDS study, glucose level on admission was not associated with altered effectiveness of thrombolysis. All of these findings suggest that glucose level is an important risk factor for morbidity and mortality after stroke. However, it is not clear whether hyperglycemia itself affects stroke outcome or reflects, as a marker, the severity of the event due to the activation of stress hormones such as cortisol or norepinephrine. Diffusion–perfusion MRI analysis supports the first hypothesis. Hyperglycemia greater than 12.1 mmol/l in patients with perfusion–diffusion mismatch, shown on diffusion-weighted imaging–perfusion-weighted imaging (DWI/PWI) MRI, was associated with higher lactate production and with reduced salvage of mismatch tissue and increased conversion of tissue “at risk” to infarcted tissue compared with patients who arrived with the value of 5.2 mmol/l . Recently, the CHANCE study of Pan et al. showed that stress hyperglycemia was associated with an increased risk of stroke in patients with a minor ischemic stroke or even transient ischemic attack .
Among the factors found to contribute to the post-acute stroke hyperglycemia  are the involvement of the insular cortex, which is known to play a role in sympathetic activation, involvement of the internal capsule, pre-existing diabetes, elevated systolic BP, and National Institutes of Health Stroke Scale (NIHSS) higher than 14 points. Interestingly, stress-related hyperglycemia after stroke was more prevalent in patients with high visceral obesity and BMI, and positive family history for diabetes. These patients can constitute a high-risk group requiring close monitoring of blood sugars and thus preventing complications .
Almost two in every three acute stroke patients have elevated blood glucose more than 6.1 mmol/l at stroke onset. Glucose level is an important risk factor for morbidity and mortality after stroke, but it is unclear whether hyperglycemia itself affects stroke outcomes or reflects the severity of the event as a marker.
Control of Hyperglycemia
The previous data raise the question of how, and especially to what extent, should post-acute stroke hyperglycemia be treated. Intensive insulin therapy administered intravenously (i.v.) and aimed at maintaining blood glucose levels at 4.5–6.1 mmol/l in the surgical intensive care set-up was found to reduce mortality by more than 40% . Similar results were documented among patients after myocardial infarction . The question remains regarding the application in acute stroke patients. The UK Glucose Insulin in Stroke Trial (GIST-UK) addressed this question . The study was conducted among 933 hyperglycemic acute stroke patients who received glucose-potassium-insulin (GKI) infusion versus placebo. In the treatment group significantly lowered glucose and BP values were documented. However, no clinical benefit was found among the treated patients. The time window for treating post-stroke hyperglycemia still remains uncertain. There are a variety of methods of insulin administration, including continuous i.v. infusion, repeated subcutaneous dosing, and i.v. infusion containing insulin and dextrose with potassium supplementation [44, 54]. The Glucose Regulation in Acute Stroke Patients Trial (GRASP) randomized 74 patients with hyperglycemia (glucose >6.1 mmol/l) within 24 hours of symptom onset to tight glucose control (3.9–6.1 mmol/l), loose glucose control (6.1–11.1 mmol/l), or normal care. The insulin is delivered as a GKI infusion. The primary outcome, the rate of hypoglycemic events, (<55 mg/dl) was 4% in usual care, 4% in loose control, and 30% in tight control; except for one all of these events were asymptomatic. The loose control group had a significantly greater proportion of deaths (25%) than the usual care group (4%); no significant difference was observed between the tight control group (13%) and the usual care group. This pilot trial was not powered to assess efficacy on clinical endpoints .
A randomized, multicenter, blinded pilot trial, Treatment of Hyperglycemia in Ischemic Stroke (THIS) , compared the use of aggressive treatment with continuous i.v. insulin, with no glucose or potassium in the insulin solution, with insulin administered subcutaneously in acute stroke patients. The aggressive-treatment group was associated with somewhat better clinical outcomes, which were not statistically significant. In the recently published INSULINFARCT study , 180 patients with acute ischemic stroke were randomized to receive either intensive insulin therapy or usual subcutaneous insulin for 24 hours. The former regimen was found to improve glucose control in the first 24 hours of stroke, but was associated with larger infarct growths as measured by MRI. The 3-month functional outcome, death, and serious adverse events were similar in both groups.
Consequently, the question arises as to what, when, and how to start therapy. Oral anti-glycemic therapy is not recommended for treatment of post-stroke hyperglycemia due to the risk of acidosis and hypoglycemia when used in acute stroke patients . Instead, insulin is the preferred treatment for a number of reasons: (1) therapeutic efficacy of insulin therapy for post-stroke hyperglycemia is proven to reduce the glucose level in acute stroke patients ; (2) i.v. or subcutaneous insulin is much easier to apply on stroke patients, especially in patients with reduced level of consciousness or in dysphagic patients ; (3) insulin promotes activity of endothelial nitric oxide ; (4) insulin has an immune-modulatory effect that can control the size of infarction and thus subsequently improve the prognosis .
American stroke guidelines recommend the use of subcutaneous insulin, because risk of hypoglycemia is lower compared to i.v. administration. Subcutaneous insulin with basal insulin regimes can usually keep appropriate normoglycemia in most patients with acute ischemic stroke . Also, monitoring and administering of the subcutaneous insulin sliding scale is easier because it is given four times a day in comparison to rigorous hourly glucose monitoring for i.v. insulin therapy. However, if the glucose level is constantly over 11.1 mmol/l, i.v. insulin infusion should be used with repeated glucose monitoring .
The precise timing and glucose value at which insulin therapy should be started and aimed remains to be determined, although earlier initiation from the first 24–48 hours after admission is practical and beneficial [38, 58].
Unfortunately, no guidelines are available regarding the duration of insulin therapy. Walters et al. found that 48 hours’ duration of insulin infusion therapy was satisfactory and treatment over 48 hours was unnecessary . On the other hand, the GIST-UK study showed that shorter 24-hour insulin duration did not have significant clinical impact . In the recent studies no association between longer duration of insulin therapy and hypoglycemia was found. The GRASP study, with 5 days’ duration of therapy as compared to THIS trial, with 24-hour duration of therapy, had almost the same incidence of hypoglycemia in the insulin treatment group. Thus, even short therapy duration can result in hypoglycemia. Therefore, 48 hours of i.v. insulin therapy seems to be the most appropriate duration .
It is worth mentioning that hypoglycemia in stroke patients generally occurs when glucose-lowering therapy is instituted and rarely occurs spontaneously. Possible benefits of controlling tight glucose levels are offset by the risk of hypoglycemia which can cause permanent neurological damage. AHA/ASA guidelines suggest that it is useful to treat hyperglycemia by maintaining blood glucose levels in a range from 7.8 to 10.0 mmol/l in all hospitalized patients [38, 65]. Also, European stroke guidelines have a similar proposal, but with a wider range from 4 to 11 mmol/l blood glucose [65. 66]. Further on, if hypoglycemia is recognized, it should be urgently treated with solution of dextrose (25 ml of 50% dextrose given in slow i.v. push) [38, 65].
For control of hyperglycemia, insulin is the preferred treatment. AHA/ASA guidelines suggest to maintain blood glucose levels in a range from 7.8 to 10.0 mmol/l (140–180 mg/dl). European stroke guidelines suggest a range from 4 to 11 mmol/l blood glucose (70–200 mg/dl).
Several animal studies demonstrated the correlation of elevated temperature and poor outcome in ischemic stroke models [67, 68]. Similar results were found in human observations. In the Copenhagen stroke study, stroke severity was correlated with hyperthermia higher than 37.5°C, while a temperature lower than 36.5°C was associated with a favorable outcome .
Other studies limited the correlation between stroke severity and hyperthermia to only the first 24 hours following stroke onset. In a prospective study temperature was recorded every 2 hours for 72 hours in 260 patients with a hemispheric ischemic stroke. Hyperthermia initiated only within the first 24 hours from stroke onset, but not afterwards, was associated with larger infarct volume and worse outcome . Hyperthermia may result from the brain infarct, but the progress of inflammatory and biochemical mechanisms associated with brain ischemia is also important. It accentuates ischemic mechanisms within the penumbra, contributing its conversion into an irreversible lesion .
It is crucial to distinguish between hyperthermia and fever. Frequently these terms are used as synonyms, but it is important to make the distinction between the two, because the origin and influence of these syndromes are different . Hyperthermia is a breakdown in thermoregulation in which there is uncontrolled heat production, poor heat dissipation, or an external heat load that does not involve a thermoregulatory set point, i.e. hyperthermia is an increase in body temperature in which pyrogenic cytokines and microbial products are not directly involved . On the other hand, fever is a reaction to pyrogens that not only cause the body’s thermoregulatory set point to increase, but also simultaneously stimulate an acute phase reaction and activate complex metabolic, immunologic, and endocrinologic systems [73, 74]. It is worth mentioning that about half of stroke patients develop fever, and those patients are far more likely to die within the first 10 days after a stroke than patients without fever after stroke . Nevertheless, high body temperature following stroke is associated with higher morbidity and mortality rates compared with patients with normal body temperature, independent of the origin of the temperature rise .
Hyperthermia and fever after stroke are different entities; both are not fully understandable from the pathophysiological point of view, but certainly any rise of body temperature may worsen stroke outcome.
The above-mentioned animal studies and human observations raised the question regarding the role of hypothermia as a treatment for acute stroke. Hypothermia was introduced more than 55 years ago as a protective measure for the brain . Mild induced hypothermia was found to improve neurological outcomes and reduce mortality following cardiac arrest due to ventricular fibrillation ; on the other hand, treatment with hypothermia aiming at 33°C within the first 8 hours after brain injury was not found to be effective . Other applications for which therapeutic hypothermia was suggested include acute encephalitis, neonatal hypoxia, and near drowning .
The use of anti-pyretics, such as acetaminophen, in high doses ranging between 3 900 and 6 000 mg daily [80, 81] caused only very mild reduction in body temperature, in the range 0.2–0.4°C respectively. The clinical benefit of this reduction is not well established. The use of external cooling aids  such as cooling blankets, cold infusions, and cold washing, aiming at a body temperature of 33°C for 48–72 hours in patients with severe middle cerebral artery (MCA) infarction was not associated with severe side-effects and was found to help control elevated intracranial pressure values in cases of severe space-occupying edema. Similar results, of decreasing acute post-ischemic cerebral edema, were found in a small pilot study of endovascular-induced hypothermia .The use of an endovascular cooling device which was inserted into the inferior vena cava was evaluated among patients with moderate to severe anterior circulation territory ischemic stroke in a randomized trial. Although no difference was found in the clinical outcome between the treatment group and the group randomized to standard medical management, the results suggest that this approach is feasible and that moderate hypothermia can be induced in patients with ischemic stroke quickly and effectively and is generally safe and well tolerated in most patients . However, the current data do not support the use of induced hypothermia for treatment of patients with acute stroke. In conclusion, despite its therapeutic potential, hypothermia as a treatment for acute stroke has been investigated in only a few very small studies. Therapeutic hypothermia is feasible in acute stroke, but owing to side-effects such as hypotension, cardiac arrhythmia, and pneumonia it is still thought of as experimental, and evidence of efficacy from clinical trials is needed . According to the 2008 ESO recommendations , at a temperature of 37.5°C or above, reducing the body temperature should be advised. The American Heart and Stroke Association  recommends that anti-pyretic agents should be administered in post-stroke febrile patients, but the effectiveness of treating either febrile or non-febrile patients with anti-pyretics is not proven.
According to newest studies, factors that affect the efficacy of hypothermia include: (1) the cooling duration; (2) the time when cooling begins; and (3) reperfusion of the occluded vessel. Hence, longer periods of hypothermia started soon after the ischemia onset in patients treated with reperfusion therapies would be expected to have the best possibility for good outcomes. Also, the intensity of cooling seems to be a less critical factor. Laboratory studies suggest that a small decrease in temperature is as protective as a larger decrease. The potential reason for this might be the increased occurrence of adverse side-effects such as non-fatal cardiac arrhythmias with lower temperature .
The optimal length of hypothermia after stroke is still unclear, but longer durations seem to be related with more consistent and long-lasting protection . Also, longer duration of hypothermia is crucial when the initiation of cooling is delayed.
The time to cooling is another significant factor, because it is well known that earlier action increases the chances of good outcome. According to experimental models, the best neuroprotective effect of hypothermia is achieved if cooling is started even before the onset of stroke. Since many patients do not present to the emergency room immediately after stroke onset, a critical problem is how long after stroke hypothermia can be applied and still be useful. The data are inconsistent. Also, it is still unclear whether longer cooling will allow longer cooling delay.
Recanalization is another factor that might increase the possibility of a beneficial outcome. Laboratory studies have shown consistent protection of hypothermia against temporary ischemia, but data from permanent ischemia models are inconsistent. These results have clear implications in clinical practice, where early recanalization may be required with mechanical or pharmacological approaches in order to achieve a beneficial effect from induced hypothermia [88, 89].
Hyperthermia within the first 24 hours from stroke onset was associated with larger infarct volume and worse outcome. The 2008 ESO Guidelines recommend reducing body temperature only if above 37.5°C.
Hypothermia is a recognized neuroprotectant in the laboratory models, showing remarkable effects. Also, it has shown benefits in clinical practice in patients with cardiac arrest and in pediatric populations experiencing hypoxic brain insults. Its role in stroke therapy has yet to be established.
In summary, hypertension, hyperglycemia, and hyperthermia are common conditions following acute stroke. All three have a major and independent impact on the severity of outcome. Occasionally, the benefit of this impact is no less than that of more “heroic” strategies such as intravenous and intra-arterial thrombolysis. Despite the lack of consensus on the data and optimal management, one should carefully monitor these three “hyper-links” and treat them appropriately.
Intermittent monitoring of neurological status, pulse, blood pressure, temperature, and oxygen saturation is recommended for 72 hours in patients with significant persisting neurological deficits
It is recommended that oxygen should be administered if the oxygen saturation falls below 95%
Regular monitoring of fluid balance and electrolytes is recommended in patients with severe stroke or swallowing problems
Normal saline (0.9%) is recommended for fluid replacement during the first 24 hours after stroke
Routine blood pressure lowering is not recommended following acute stroke
Cautious blood pressure lowering is recommended in patients with extremely high blood pressures (> 220/120 mmHg) on repeated measurements, with severe cardiac failure, aortic dissection, or hypertensive encephalopathy
It is recommended that abrupt blood pressure lowering be avoided
It is recommended that low blood pressure secondary to hypovolemia or associated with neurological deterioration in acute stroke should be treated with volume expanders
Monitoring serum glucose levels is recommended
Treatment of serum glucose levels >180 mg/dl (>10 mmol/l) with insulin titration is recommended
It is recommended that severe hypoglycemia (< 50 mg/dl [< 2.8 mmol/l]) should be treated with intravenous dextrose or infusion of 10–20% glucose
It is recommended that the presence of pyrexia (temperature >37.5 °C) should prompt a search for concurrent infection
Treatment of pyrexia (temperature >37.5 °C) with paracetamol and fanning is recommended
Antibiotic prophylaxis is not recommended in immunocompetent patients
Management of Post-Stroke Complications
Stroke is a major cause of long-term physical, cognitive, emotional, and social disability. In addition to the neurological impairment appearing in the acute phase, there are infrequently late complications which are often neglected. These complications have a great impact on the quality of life, outcome, and chances of rehabilitation and may include post-stroke epilepsy, dementia, depression, and fatigue. Other complications, such as infections, are dealt with in Chapter 23. Box 22.2 gives an overview of the recommendations of the ESO for the prevention and management of complications .
It is recommended that infections after stroke should be treated with appropriate antibiotics
Prophylactic administration of antibiotics is not recommended, and levofloxacin can be detrimental in acute stroke patients
Early rehydration and graded compression stockings are recommended to reduce the incidence of venous thromboembolism
Early mobilization is recommended to prevent complications such as aspiration pneumonia, DVT, and pressure ulcers
It is recommended that low-dose subcutaneous heparin or low molecular weight heparins should be considered for patients at high risk of DVT or pulmonary embolism
Administration of anti-convulsants is recommended to prevent recurrent post-stroke seizures. Prophylactic administration of anti-convulsants to patients with recent stroke who have not had seizures is not recommended
An assessment of risk of falls is recommended for every stroke patient
Calcium/vitamin D supplements are recommended in stroke patients at risk of falls. Bisphosphonates (alendronate, etidronate, and risedronate) are recommended in women with previous fractures
In stroke patients with urinary incontinence, specialist assessment and management are recommended
Swallowing assessment is recommended, but there are insufficient data to recommend a specific approach for treatment
Oral dietary supplements are only recommended for non-dysphagic stroke patients who are malnourished
Early commencement of nasogastric (NG) feeding (within 48 hours) is recommended in stroke patients with impaired swallowing
It is recommended that percutaneous enteral gastrostomy (PEG) feeding should not be considered in stroke patients in the first 2 weeks
Epilepsy is one of the most common serious neurological disorders and is associated with numerous social and psychological consequences. Stroke is the most commonly identified etiology of secondary epilepsy and accounts for 30% of newly diagnosed seizures in patients older than 60 years . Although recognized as a major cause of epilepsy in the elderly, many questions still arise regarding the epidemiology, treatment, and outcome of post-stroke seizures.
The common definition of epilepsy includes at least two seizures with a time interval of at least 24 hours between the episodes. The current clinical classification of post-stroke seizures is made according to the period between the stroke and the first epileptic episode. A post-stroke seizure is defined as early if it occurs in the first 2 weeks after the stroke. A seizure occurring later than this is defined as late .
The estimated rate of early post-ischemic stroke seizures ranges from 2% to 33% and that of late seizures varies from 3% to 67% [92–100]. The wide range is due to the different methodologies, terminologies, and sizes of the populations in the different studies. The overall rate of post-stroke epilepsy, as previously defined as at least two episodes, is 3–4% and is higher in patients who have had a late seizure .
In an observational study among 1 428 patients after stroke , 51 patients (3.6%) developed epilepsy. Post-stroke epilepsy was found to be more common among patients with hemorrhagic strokes, venous infarctions, and localization in the right hemisphere and MCA territory. The SASS (Seizures After Stroke Study) was a prospective multicenter study held among 1 897 patients after an ischemic or a hemorrhagic stroke . In that study 14% of the patients with ischemic stroke and 20% of patients with hemorrhagic stroke had seizures during the first year; a second episode, required to establish epilepsy, was found in 2.5% of the patients. Most of the patients with post-stroke epilepsy have simple partial seizures, while complex partial seizures are relatively rare. The risk of status epilepticus varies from 0.14% to 13%. It should be emphasized that it is not always clear whether the patient has had a seizure; seizures in the elderly are sometimes difficult to diagnose and may present as acute confusion, behavioral changes, or syncope of unknown origin .
Post-stroke epilepsy is defined as at least two episodes of seizures. The overall rate is 3–4% of stroke patients.
Other predictors for post-stroke seizures found in various studies are cortical location, large infarct, evaluated clinically or radiologically, intracerebral hemorrhage, and cardiac emboli, most probably due to the tendency of the last to involve the cortex . Post-stroke seizures are also more common among patients with pre-existing dementia evaluated using the validated IQCODE questionnaire (risk ratio of 4.66, CI 1.34–16.21). A recent cohort study found major stroke and sinus thrombosis as the two major predictors for post-stroke epilepsy . In that study, conventional vascular risk factors were not associated with the occurrence of post-stroke seizures. Patients in high-risk populations should be advised to avoid factors increasing the risk of seizures, such as certain drugs . In a retrospective study the presence of chronic obstructive pulmonary disease (COPD) was found to be an independent risk factor for the development of seizures in stroke patients .
There are a number of presumed causes for early-onset seizures after strokes. An increase in intracellular Na+ and Ca2+ with a consequential lower threshold for depolarization, increased excitatory activity mediated by the release of glutamate from the hypoxic tissue, metabolic dysfunction, global hypoperfusion, and hyperperfusion injury have all been postulated as presumed etiologies [101, 105]. Late seizures are due to the development of tissue gliosis and neuronal damage in the infarct area . An interesting question is whether post-stroke seizures worsen the outcome of patients after stroke. A cortical cerebral infarction disability was found to be greater in patients with seizures; on the other hand, in patients with cortical hemorrhage disability was found to be less .
The attending physician is required to deal with two important questions, the first being whether to start treatment after the first seizure episode and the second being which anti-epileptic drug to prefer. According to the common clinical approach, treatment should be initiated only after the second episode. Observational studies suggest that isolated early seizures after stroke do not require treatment [94, 95]. Beginning treatment after early-onset seizures has not been associated with reduction of recurrent seizures after discontinuing the medication . However, according to new ESO guidelines for the management of post-stroke seizures and epilepsy (2017), secondary anti-epileptic drugs prophylaxis needs to be considered, due to high incidence of epileptic seizure recurrence after one post-stroke unprovoked seizure. Unfortunately, due to very low evidence, based on few randomized controlled trials and observational studies, ESO guidelines only give some weak recommendations on prevention of occurrence and recurrence of post-stroke seizures. Adequately powered randomized controlled studies are necessary to assess interventions for post-stroke seizure and epilepsy management .
At this stage there are no evidence-based studies to recommend one drug over the others. It is best to avoid the old drugs, especially phenytoin, because of their pharmacokinetic profile and interactions with anti-coagulants and salicylates . “New-generation” drugs (lamotrigine, gabapentin, and levetiracetam, etc.) in low doses would be a reasonable option because of their efficacy, improved safety profile, and fewer interactions with other drugs compared with first-generation drugs . A single study has found gabapentin to be a safe and effective treatment; however, this recommendation should be taken with caution since the study had no control group . In a prospective study comparing lamotrigine versus carbamazepine in 64 patients with post-stroke epilepsy, lamotrigine was found to be significantly better tolerated and with a trend to be also more efficacious (p = 0.06) .
There is no evidence to prefer one anti-epileptic drug over the others, but it is advised to avoid phenytoin because of interactions with anti-coagulants and salicylates.
Post-stroke depression is considered to be the most frequent and important neuropsychiatric consequence of stroke and has a major impact on functional recovery, cognition, and even survival.
The incidence of post-stroke depression ranges in various studies between 18% and 61%. Once again, the large variation in frequencies is due to methodological differences, including the point in time at which patients were assessed relative to the stroke onset and the different instruments and criteria for diagnosing depression that were used in the different studies.
A systematic review of collected data from 51 observational studies conducted between 1977 and 2002 found that the frequency of post-stroke depression is 33% (95% CI 29–36) and that the depression resolves spontaneously within several months of onset in most of the patients . The Italian multicenter observational study of post-stroke depression (DESTRO) assessed 1 064 patients with ischemic or hemorrhagic stroke in the first 9 months after the event . Patients with depression were followed for 2 years. Post-stroke depression was detected in 36% of the patients, most of whom had minor depression with dysthymia, rather than major depression, and adaptation disorder. Although no correlation between post-stroke depression and mortality was found in the DESTRO study, an Australian study  found that among stroke patients in rehabilitation, those who were depressed were eight times more likely to have died by 15-month follow-up than those who were not depressed. In a recent Chinese multicentered prospective cohort study the diagnosis of post-stroke depression was associated with an increased risk of recurrent stroke at 1 year in a multivariate model (OR 1.49; 95% CI 1.03–2.15) .
The potential etiology for post-stroke depression  includes neuroanatomical mechanisms such as disruption of monoaminergic pathways and depletion of cortical biogenic amines, especially in the case of lesions in the left frontal and left basal ganglia territories , and psychological mechanisms such as the difficulty in adjusting to the new limitations and requirements of the disease. In a systematic review of 26 studies regarding the correlation of left hemispheric stroke and the risk of post-stroke depression, no significant correlation was found . Differences in the measurement of depression, study design, and presentations of results may also have contributed to the heterogeneity of the findings. Other risk factors for post-stroke depression include female gender, severe physical disability, previous depression, and history of psychiatric and emotional liability during the first days after stroke. Some studies have found aphasia as a risk factor, while others have not obtained similar results . Dementia was also found to be an important predictor for the development of post-stroke depression .
The frequency of post-stroke depression is 33% and it resolves spontaneously within several months of onset in most patients.
According to a recent meta-analysis which included 36 studies and identified risk factors for post-stroke depression occurrence, a history of mental illness was the highest ranking modifiable risk factor. Other risk factors were age (<70 years), neuroticism, female gender, family history, level of disability, and stroke severity. Social support was a protective factor for post-stroke depression .
The treating physician should be aware of the diagnosis of depression in stroke survivors since it may be hindered by a number of conditions, including aphasia, agnosia, apraxia, and memory disturbances. The differential diagnosis of post-stroke depression includes anosognosia, apathy, fatigue, and disprosody . Despite some encouraging data regarding the prophylactic use of anti-depressants in post-stroke patients, there is still insufficient randomized evidence to support this approach in routine post-stroke management . A single recent double-blind placebo-controlled study evaluated the administration of escitalopram in a population of non-depressed patients following stroke . Patients who received placebo were significantly more likely to develop depression than ones who received escitalopram after 12 months follow-up. Problem-solving therapy did not achieve significant results over placebo. A retrospective study held among 870 post-stroke patients showed that selective serotonin reuptake inhibitor (SSRI) treatment was associated with longer survival even though depression diagnosis was associated with greater risk of mortality .
According to the ESO 2008 recommendations  anti-depressant drugs such as SSRIs and heterocyclics can improve mood after stroke, but there is less evidence that these agents can effect full remission of a major depressive episode or prevent depression. SSRIs are better tolerated than heterocyclics. There is no good evidence to recommend psychotherapy for treatment or prevention of post-stroke depression, although such therapy can elevate mood.
In spite of growing information, many questions still surround various aspects of post-stroke depression, including the development of standardized measure of depression, the optimal time after stroke onset to screen for post-stroke depression, the creation of predictors for post-stroke depression, and identifying the appropriate management. However, a recent meta-analysis study conducted by Sun et al., after weighing the acceptability and efficacy of 10 anti-depressants and comparing them to placebo, showed that paroxetine might be the best choice when starting acute therapy for post-stroke depression, and fluoxetine might be the worst choice .
Anti-depressant drugs can improve mood after stroke, but there is less evidence that these agents can be effective in a major depressive episode or prevention.
Stroke is an important risk factor for dementia and cognitive decline. According to the NINDS-AIREN criteria, in order to make the diagnosis of post-stroke dementia (PSD) the patient has to be demented, with either historical, clinical, or radiological evidence of cerebrovascular disease and the two disorders must be reasonably related . On the other hand, according to the fourth edition of DSM-IV , vascular dementia is diagnosed by the development of multiple cognitive deficits manifested by memory impairment and at least one of the following cognitive disturbances: aphasia, apraxia, agnosia, and disturbance in executive functioning with the presence of focal neurological signs and symptoms or laboratory evidence indicative of cerebrovascular disease that is judged to be etiologically related to the disturbance. The deficits should not occur exclusively during the course of an episode of delirium.
Despite the lack of accurate data due to poor definition of the disorder, the use of different tools, and diagnostic difficulties in distinguishing between PSD and other types of dementia, PSD is considered to be the second most common type of dementia. Since several studies used different tools for the diagnosis of PSD and there were also differences in the methodologies and study populations, the incidence varies in the different studies from 8% to 30%. One study, done among a population of elderly demented patients, demonstrated that the frequency of dementia was found to depend upon the diagnostic criteria used . For instance, using the NINDS-AIREN criteria only 14% of the patients were diagnosed with PSD, compared to 76% using the DSM-IV as a diagnostic tool. Interestingly there are also noticeable differences in the incidence rates between countries; an almost 3-fold difference in the age-standardized incidence ratios (SIR) of PSD rates between Germany and the Netherlands was demonstrated (1.23 and 0.42, respectively) , indicating that geographical variation is still present after taking into account the countries’ differential age distributions. It is unclear whether these differences are due to genetic or environmental factors since, as in the previous trials mentioned, there were methodological differences between the studies. A recent comprehensive review on PSD proposed a simple definition: PSD accounts for any dementia following stroke in temporal relation , which is a good proposal to standardize the methodology.
Despite the conflicting data the overall estimated frequency of dementia in post-stroke patients is about 28% and the fact that stroke is a major risk factor for dementia is well established . The mechanisms of PSD [132, 133] consist of large-vessel disease, including multi-infarcts or single infarcts in a strategic area such as the thalamus, hippocampus, basal forebrain, or the angular gyrus, or small-vessel disease such as lacunes or leukoaraiosis. Other mechanisms include hypoperfusion, hypoxic-ischemic disorders, and shared pathogenic pathways with degenerative dementia, especially Alzheimer type.
Risk factors for PSD include large and left-sided infarcts, bilateral infarcts, frontal lobe infarcts, large MCA infarcts, and previous strokes. Diabetes, hyperlipidemia, and atrial fibrillation were also found as predictors for the development of PSD [133–135]. Silent brain infarcts demonstrated on CT, however, were not found to predict the development of PSD in one prospective study , while in another, higher grades of white matter findings on MRI were associated with impaired cognitive function . Since it has also been shown in that study that the extent of white matter lesions is related to the BP level, even in normotensive patients, and since these lesions are correlated with the risk of PSD, it would be reasonable to assume that lowering BP would lower the risk of PSD. Abnormal EEG performed close to the ischemic stroke appears to be an indicator of subsequent PSD in a prospective study done among 199 patients, probably because it indicates cortical involvement .
The borders between dementia of the neurodegenerative type and vascular dementia are nowadays less visible and both types of dementia include many similar risk factors and clinical and pathological characteristics. It is suggested that cerebrovascular disease may play an important role in the presence and severity of Alzheimer’s disease .
There is no evidence-based treatment for PSD. In a meta-analysis of randomized controlled trials cholinesterase inhibitors, which are administered for the treatment of degenerative-type dementia, were found to produce only small benefits in cognition of uncertain clinical significance in patients with mild to moderate vascular dementia. There are insufficient data to recommend the use of these agents in PSD .
PSD prevention can be achieved by prevention of recurrent stroke. As treatment strategies to inhibit the development and mitigate the course of PSD, lowering of BP, neuroprotective drugs, statins, and anti-inflammatory agents have all been studied, but unfortunately without convincing evidence of efficacy. Physical activity, lifestyle interventions, and cognitive training have been recently tested, but large controlled trials are still missing .
The frequency of dementia (PSD) in post-stroke patients is about 28%. There is no evidence-based treatment for PSD.