Management of patients with hypertensive crises.
Hypertensive crises: urgencies and emergencies
Definitions and probable causes of hypertensive crisis
Hypertensive urgencies can be defined as severe elevations in BP that do not exhibit evidence of target-organ (cardiovascular, renal, CNS) dysfunction or damage. Urgencies can be managed by the administration of oral medications, most often in the emergency department (ED), and follow-up on an outpatient basis. Hypertensive emergencies are severe elevations in systolic and diastolic BP associated with acute target-organ damage that require immediate management in a hospital setting.
Hypertensive crises encompass a spectrum of clinical situations that have in common blood pressure (BP) that is elevated, and progressive or impending target organ damage. Most hypertensive urgencies or emergencies are preventable and are the result of inadequate treatment of mild-to-moderate hypertension or nonadherence to antihypertensive therapy.
Traditionally, hypertensive crises have been divided into emergencies and urgencies. Hypertensive emergencies are severe elevations in blood pressure (BP) that are complicated by evidence of progressive target organ dysfunction, and will require immediate BP reduction (not necessarily to normal ranges) to prevent or limit target organ damage. Examples include: hypertensive encephalopathy, intracranial hemorrhage, unstable angina pectoris, or acute myocardial infarction, acute left ventricular failure with pulmonary edema, dissecting aneurysm, or eclampsia. While the level of BP at the time of presentation is usually very high (greater than 180/120 mm Hg), keep in mind that it is not the degree of BP elevation, but rather the clinical status of the patient that defines a hypertensive emergency. For example, a BP of 160/100 mm Hg in a 60-year-old patient who presents with acute pulmonary edema represents a true hypertensive emergency.
Hypertensive urgencies are severe elevations of BP but without evidence of progressive target organ dysfunction and would be better defined as severe elevations in BP without acute, progressive target organ damage. A traditional term “urgency” has led to aggressive and often excessive treatment of the majority of patients who present to Emergency Departments (ED) with severe hypertension. While these patients may present with levels of BP similar to the hypertensive emergency, and may have evidence of target organ involvement, they do not display evidence of ongoing progressive target organ damage. Most of these patients are, in fact, nonadherent to drug therapy or are inadequately treated hypertensive patients and often present to the ED for other reasons. Patients with severe elevations of BP can be managed in the ED with oral agents and appropriate follow-up within 24 hours to several days depending upon the individual characteristics of the patient. It is the correct differentiation of these two forms of hypertensive crises, however, that presents the greatest challenge to the physician.
A brief but thorough history should address the duration as well as the severity of hypertension, all current medications including prescription and nonprescription drugs and, of particular importance, the use of recreational drugs. A history of other comorbid conditions and prior cardiovascular or renal disease is essential to the initial evaluation. Direct questioning regarding the level of compliance with current antihypertensive medications may establish inadequacy of treatment or frank noncompliance.
Frequent or continuous monitoring of BP should be established. Look for historical information suggestive of neurologic, cardiovascular, and/or renal symptoms. Check for specific manifestations such as headache, seizures, chest pain, dyspnea, and edema. The clinical characteristics of a hypertensive emergency are listed here. The level of BP alone does not determine a hypertensive emergency; rather, it is the degree of target organ involvement that will determine the rapidity with which BP should be reduced to a safer level to prevent or limit target organ damage. Initial therapy will be for a presumptive diagnosis based on the information available during the initial triage evaluation.
The attached algorithm (See algorithm) can help the clinician identify those patients who meet the criteria of a hypertensive emergency that requires immediate admission to an ICU for continuous monitoring of BP and initiation of parenteral antihypertensive therapy. For patients with uncontrolled hypertension (urgency), evidence of target organ damage may or may not be present but these patients do not demonstrate any evidence of deterioration of target organ function. They can be observed for several hours in the ED during which time their oral medications can be resumed, if discontinued, or if untreated, an oral regimen can be initiated. On occasion, increasing presently inadequate dosages of medication may be appropriate. Appropriate outpatient follow-up can then be arranged within 24 hours to several days as needed, and if no prior evaluation has been performed on this patient for hypertension, an outpatient appointment should be established. Failure to follow-up on this large group of patients is a missed opportunity from the standpoint both of keeping patients in the healthcare system, and establishing optimal BP.
The physical examination should begin with an assessment of BP, with an appropriate-size cuff in both upper extremities and in a lower extremity if peripheral pulses are markedly reduced. Brachial, femoral, and carotid pulses should be assessed. A careful cardiovascular examination as well as a thorough neurologic examination, including mental status, should be conducted. This assessment will aid in establishing the degree of involvement of affected target organs and should provide clues to the possible existence of a secondary form of hypertension, such as renovascular hypertension. If a secondary cause of hypertension is suspected, appropriate blood and urine samples should be obtained before aggressive therapy is initiated. A careful funduscopic examination should be performed to detect the presence of hemorrhages, exudates, and/or papilledema.
Initial Laboratory Studies
Initial laboratory studies should be limited and rapidly expedited. A urinalysis with microscopic examination of the urinary sediment, an immediate chemistry panel, and an electrocardiogram should be obtained. The urinalysis may reveal significant proteinuria, red blood cells, and/or cellular casts. Cellular casts are suggestive of renal parenchyma disease. Electrolyte abnormalities, particularly hypokalemia or hypomagnesemia, increase the risk of cardiac arrhythmias, and the chemistry panel will also provide evidence of renal and/or hepatic dysfunction. The electrocardiogram should identify evidence of coronary ischemia and/or left ventricular hypertrophy and may reveal pulse deficits, raising the question of aortic dissection. When the clinical examination suggests cerebrovascular ischemia or hemorrhage, or if the patient is comatose, a computed tomographic scan of the head should be immediately obtained.
Initial Treatment of the Hypertensive Emergency
The initial goal for BP reduction is not to obtain a normal BP but rather to achieve a progressive, controlled reduction in BP to minimize the risk of hypoperfusion in cerebral, coronary, and renovascular beds. Under normal conditions, blood flow to these organs remains relatively constant despite wide fluctuations in BP. In the presence of severe hypertension, the autoregulatory range is shifted upward so that higher levels of BP are tolerated, but organ circulation may be put at risk with sudden reductions in BP. As an example, studies on the autoregulation of cerebral blood flow suggest that the lower limit of autoregulation is about 25% below the resting mean arterial pressure in normotensive subjects and in those with uncomplicated essential hypertension. These observations have led to the suggestion that initial reduction in mean arterial pressure should not exceed 20% to 25% below the pretreatment BP. As an alternative, mean arterial pressure can be reduced within the first 30 to 60 minutes to 110 to 115 mm Hg. If this level of BP is well tolerated and the patient is clinically stable, further gradual reductions toward a normal BP can be implemented over the next 24 hours. Excessively rapid reductions in BP have been associated with acute deterioration in renal function, ischemic cardiac or cerebral events, and occasional retinal arterial occlusion and acute blindness.
A significant exception to the above recommendations should be recognized for patients with ischemic stroke, with the awareness that cerebral autoregulation is disrupted in ischemic tissue. There is no clear evidence from clinical trials to support the use of antihypertensive treatment during an acute stroke in the absence of other concurrent disorders such as aortic dissection or heart failure. Antihypertensive treatment may adversely affect cerebral autoregulation in acute stroke. Hypertension associated with an acute ischemic stroke spontaneously decreases to pre-stroke levels within several days.
How Urgent is Urgent Hypertension?
Historically, most patients seen in the ED with severe hypertension did not meet the criteria for hypertensive emergency and were therefore classified as having hypertensive urgencies. Most were treated aggressively in the ED and many were, in fact, admitted to the hospital for control of BP. The important caveat is that elevated blood pressure alone rarely requires emergency therapy. Initial triage should identify those patients who have an elevated BP without any evidence of significant target organ damage or any other impending cardiovascular events, and these patients clearly represent the majority of those seen in the ED. They are often asymptomatic and can be observed for a brief period in the ED to initiate or resume medication in the noncompliant patient or to increase dosages for those being inadequately treated. Follow-up can be arranged within several days in the outpatient department.
The occasional patient who presents to the ED with uncontrolled hypertension and symptoms such as headache, shortness of breath, or epistaxis, may benefit from observation in the ED over a period of several hours and/or an increase in current medications or added medication to further lower BP under observation and monitoring of current symptoms. When clinically stable, the patient may safely be sent home with oral agents and arrangements for follow-up. There are several oral agents available that can provide rapid response in blood pressure within one to several hours. These include agents such as the short-acting ACE inhibitor, captopril, clonidine, Labetalol, or in selected patients the alpha-adrenergic blocking agent prazosin. Some of the pharmacologic characteristics of these oral agents are listed below.
In either case, to discharge the patient from an ED without a confirmed follow-up appointment represents a missed opportunity to get that patient back into treatment for optimal control of BP, which should be a management goal. There is little justification today to admit patients with hypertensive urgency or high BP to a hospital for further evaluation and management when these issues can be efficiently and cost-effectively addressed in the outpatient setting.
Captopril, an angiotensin-converting enzyme inhibitor, is well tolerated and can effectively reduce BP in a hypertensive urgency. Given by mouth, captopril is usually effective within 15 to 30 minutes and may be repeated in 1 to 2 hours, depending on the response. The drug has been administered sublingually, in which case the onset of action is within 10 to 20 minutes, with a maximal effect reached within 1 hour. Responsiveness to captopril can be enhanced by the coadministration of a loop diuretic such as furosemide. Administration may lead to acute renal failure in patients with high-grade bilateral renal artery stenosis, and some reflex tachycardia may be observed.
Clonidine is a centrally acting alpha-adrenergic agonist with onset of action 30 to 60 minutes after oral administration, and maximal effects are usually seen within 2 to 4 hours. This agent is most commonly administered as a loading dose of 0.1 or 0.2 mg followed by 0.1 mg hourly for several hours until an appropriate BP level is attained. Evidence suggests that a comparable response may be seen with a single 0.2 mg dose.5 The most common adverse effect in the acute setting is drowsiness, affecting up to 45% of patients. Clonidine may be a poor choice when monitoring of mental status is important. Dry mouth is a common complaint, and light-headedness is occasionally observed.
Labetalol, a combined alpha- and beta-adrenergic blocking agent, can be effectively administered orally in a dose of 200 to 400 mg with BP response observed within 2 to 3 hours. However, the onset of effect may be observed within 1 to 2 hours. Like other beta-blocking agents, labetalol has the potential to induce heart block and to worsen the symptoms of bronchospasm in the asthmatic patient. Caution must be observed in patients who have more than first-degree heart block, symptomatic bradycardia, or congestive heart failure.
Prazosin is an alpha-adrenergic blocking agent that can have limited benefit in the early management of a patient with a pheochromocytoma. Side effects include first-dose syncope, palpitations, tachycardia, and orthostatic hypotension.
Parenteral Agents for Hypertensive Emergencies
Parental therapy may be initiated in the ED if suitable supervision and monitoring of BP can be provided. More appropriately, the patient should be admitted to an ICU where monitoring of BP is available. A number of parenteral agents are effective in treating hypertensive emergencies (Table).
Labetalol is effective when administered as 20- to 40-mg bolus intravenous (IV) injections and can provide a step-wise, controlled reduction in BP to a predetermined goal. Once the goal BP is achieved, injections are stopped, and the long duration of action facilitates conversion to effective oral therapy. A continuous infusion of labetalol at 2 mg/min offers an alternative method of administration and is also associated with a gradual yet controlled reduction in BP. Since the beta-blocking effects predominate with this agent, bradycardia or heart block may be observed in patients with intrinsic heart disease.
Sodium nitroprusside has an extremely rapid onset of action, within seconds of initiating an infusion, and a rapid offset of effect within 1 to 2 minutes, which necessitates constant supervision of BP during administration. This agent is particularly effective in reducing preload and afterload in patients with impaired left ventricular function, and a carefully titrated infusion can achieve any desired goal BP. Nitroprusside does not cause sedation or somnolence but is rapidly degraded by light, requiring periodic exchange of solutions.
In patients with significant renal impairment, accumulation of a major metabolite, thiocyanate, may occur over several days with toxic effects. In the presence of poor tissue perfusion and depressed hepatic function, an intermediate metabolite in the form of cyanide can accumulate and occasionally lead to cyanide poisoning.
Nicardipine is an IV form of the dihydropyridine calcium antagonist and is effective in a high percentage of hypertensive emergencies, particularly at higher infusion rates. Its growing popularity can be attributed to its ease of administration. Infusion rates can be increased by 2.5 mg/hr at intervals of 15 to 20 minutes until the maximal recommended dosage of 15 mg/hr is obtained or until the desired reduction in BP is achieved. Dosing of nicardipine is not dependent on body weight. Nicardipine has been shown to reduce both cerebral and coronary ischemia although headache, nausea, and vomiting may occasionally be observed.
Nitroglycerin may be of particular benefit in hypertensive emergencies with coexisting coronary ischemia. Since this agent dilates collateral coronary vessels and, like nitroprusside, has a rapid onset and offset of effect, its use requires close nursing supervision. Unfortunately, at low infusion rates, nitroglycerin has its primary effect on capacitance vessels in which much higher infusion rates are required to effect arteriolar vasodilitation. Infusion rates may be increased at 3- to 5-minute intervals until the desired effect is achieved. Nitroglycerin may be particularly useful in patients with severe coronary ischemia whose BPs are only modestly elevated or in patients with acute hypertension following postcoronary artery bypass surgery. Tolerance to IV nitroglycerin may be observed within 24 to 48 hours of instituting an infusion.
Fenoldapam is a selective, peripheral dopamine1-receptor agonist that induces systemic vasodilation, particularly in the renal circulation.10 This agent also has effects on renal proximal and distal tubules. Fenoldapam does not bind to dopamine2 receptors or beta-adrenergic receptors; it has no alpha-adrenergic agonist effects but is an alpha1 antagonist. This drug does not cross the blood/brain barrier.
Fenoldapam's unique effects on the kidney provide increased urine flow rate, sodium and potassium excretion, and improved creatinine clearance, making this agent particularly attractive in hypertensive emergencies associated with significant renal impairment.
Fenoldapam provides clinical effects similar to those of nitroprusside in improving cardiac hemodynamics in patients with acute congestive heart failure. Onset of clinical effect is seen within 5 minutes, and effects tend to dissipate within 30 minutes of discontinuing the infusion. The most common side effects include headache, flushing, tachycardia, and dizziness. A dose-related increase in intraocular pressure has been observed in normotensive and hypertensive patients. Inactive metabolites are eliminated in the urine, and no dosage adjustments are required for patients with renal or hepatic impairment.
Hydralazine finds limited use today in pregnant women with preeclampsia. Five to 20 mg may be administered IV as a bolus injection, and may be repeated. The major advantage is this agent's ability to improve uterine blood flow. Hydralazine is contraindicated in patients who have coronary atherosclerosis, as administration is associated with reflex tachycardia and sodium and water retention, together with intense flushing. Headache and increased intracranial pressure have also been observed.
Other Parenteral Agents
Enalaprilat is administered in an IV dose of 1.25 mg and may be repeated at 6-hour intervals. Onset of action is within 30 minutes. Response to enalaprilat in hypertensive emergencies is unpredictable, in part because of variable degrees of plasma volume expansion. This agent may be particularly suitable in hypertensive emergencies associated with congestive heart failure or with high plasma angiotensin II concentrations.
Esmolol is an ultra-short-acting beta-adrenergic blocker that is administered intravenously. Onset of effect is seen within 1 to 5 minutes, with a rapid offset of effect within 15 to 30 minutes after discontinuation. Esmolol may be administered as a 500-µg/kg bolus injection, which may be repeated after 5 minutes. Alternatively, an infusion of 50 to 100 µg/kg/min may be initiated and increased to 300 µg/kg/min as needed. Adverse effects include increased heart block, precipitation of congestive heart failure, and bronchial spasm.
Phentolamine is a nonselective alpha-adrenergic blocking agent that is still used when excess catecholamine states, such as pheochromocytoma, are suspected. It is useful as a diagnostic agent when administered as a bolus injection of 5 to 10 mg in patients with suspected pheochromocytoma. Acute BP lowering will be seen within several minutes and may last 10 to 30 minutes. Tachycardia is common and on occasion may precipitate myocardial ischemia. Nitroprusside and labetalol are more easily titrated in the management of hypertensive emergencies associated with high circulating levels of catecholamines, and therefore phentolamine is rarely used therapeutically today.
Diazoxide is rarely used any longer in the treatment of hypertensive emergencies. Although a potent vasodilator, large doses of 300 mg were often associated with severe hypotension. Smaller miniboluses of 50 mg administered at 10- to 15-minute intervals can provide controlled reduction of BP but are usually associated with reflex tachycardia, hyperglycemia, hyperuricemia, and sodium and water retention. In view of these side effects, diazoxide offers little advantage over several other agents that have more acceptable adverse-effect profiles.
Hypertensive emergencies can carry a poor prognosis unless the BP is reduced quickly, whereas hypertensive urgencies pose less immediate danger. Hypertensive emergencies are considered to be clinical circumstances in which the BP should be reduced immediately (within a matter of a few hours), whereas in hypertensive urgencies, BP reduction can be accomplished gradually (within several hours or even days). There is no arbitrary level of BP that separates hypertensive emergencies from urgencies. The presence of acute target-organ damage is more indicative of an emergency than an urgency. In either circumstance, the complications of hypertensive crisis are reversible, if the treatment is provided efficiently.3
Any type of hypertensive disorder may be complicated by the development of hypertensive crisis. The important proviso is that it is the level of the BP—rather than the cause of hypertension—that is the main determinant of hypertensive crisis.4 (In select forms of hypertensive crises, however, the rapidity with which the BP rises seems to be more relevant than the actual level of the BP.) In most clinical circumstances, immediate BP reduction is indicated, not because of its absolute level but because the coexisting complications may make any level of hypertension risky. The risk-benefit ratio of immediate therapy in many forms of hypertensive crisis is not well-established due to the paucity of controlled clinical trials.
The most critical decision in the management of hypertensive emergencies is to assess the patient's clinical state and to ascertain whether the patient's condition dictates emergency management. The absolute indications for treatment and optimal management depend on the underlying and concomitant conditions. A patient with a true hypertensive crisis should be treated in an ICU.
The choice of oral versus parenteral drug therapy depends on the urgency of the situation, as well as on the patient's general condition. The level to which the BP should be lowered varies with the type of hypertensive crisis and should be individualized. The choice of parenteral drug is governed by the clinical manifestations and concomitant medical problems associated with hypertensive crisis. There is no predetermined level for the goal of therapy.
Complications of therapy, mainly hypotension and ischemic brain damage, can occur in patients given multiple potent antihypertensive drugs in large doses without adequate monitoring.5 Such complications should be minimized by gentle lowering of BP, careful surveillance, and individualization of therapy.
A relatively asymptomatic patient who presents with severe hypertension, (a diastolic BP of 130-140 mm Hg), need not be treated with parenteral drugs. These patients should be managed on an individual basis, and the usual course would be to intensify or alter the previous antihypertensive therapy.
Once the hypertensive crisis is resolved and the patient's clinical condition is stable, the physician should look into factors that might have contributed to the dangerous elevation of BP (nonadherence to prescribed therapy or the presence and/or progression of a secondary form of hypertension such as a renal artery stenosis or renal failure, for example). The physician should discuss long-range and periodic outpatient follow-up plans with the patient, as close follow-up is extremely important.
Accelerated and malignant hypertension Accelerated hypertension is clinically identified by severe retinopathy (without papilledema), exudates, hemorrhages, arteriolar narrowing, and spasm. Malignant hypertension is considered a deterioration of the accelerated form and is distinguished by papilledema. Both the accelerated and malignant forms of hypertension are associated with severe vascular injury to the kidney and other target organs. The diagnosis of accelerated or malignant hypertension is best made clinically on the basis of history and clinical examination. Simple investigations such as chest radiography, ECG, urinalysis, and laboratory tests to determine blood count, serum nitrogen level, creatinine clearance, and electrolyte concentrations are generally sufficient.
The BP level in malignant hypertension is often quite high, with diastolic levels in the range of 130 to 140 mm Hg (stage 2 hypertension), but the degree of BP elevation is an unreliable diagnostic feature (see Table 2). Rather, it is the degree and extent of vascular injury that determines the clinical manifestations. Severe headache, with or without coexisting encephalopathy, is one symptom. Weight loss may occur in some patients with malignant hypertension as a result of initial natriuresis.
Some patients with malignant hypertension report visual problems ranging from blurring to blindness; drowsiness and altered mental status are also common. Worsening symptoms may indicate progression to encephalopathy or other cerebral complications, such as stroke. Congestive heart failure can be a feature of malignant hypertension, either as a consequence of left ventricular dysfunction, or due to volume retention from concomitant renal failure. Azotemia, a frequent feature of malignant hypertension, may be associated with proteinuria. Renal function deteriorates rapidly in many patients without appropriate therapeutic intervention.
Patients with accelerated/malignant hypertension should ideally be treated in the hospital, since the goal of management is not only to lower the BP but to stabilize it, reverse the target-organ damage, and exclude any reversible causes. Preferably, these patients should be treated in an ICU; but in the absence of significant target-organ dysfunction, they can be managed safely on the wards under supervision.
Although IV therapy is widely used in the initial treatment of malignant hypertension, various oral therapies can also be used successfully. ACE inhibitors, minoxidil, clonidine, prazosin, labetalol, and nifedipine have all been used for initial treatment of malignant hypertension, but there are no controlled prospective studies to offer precise guidelines on the preferred therapeutic options. The choice between oral and parenteral therapy depends on the monitoring facilities, condition of the patient, and coexisting complications. Once the BP is brought to safe levels, appropriate oral therapy must be initiated on the basis of the patient's renal, cardiac, and neurologic status.
Hypertensive encephalopathy is a deadly complication of severe hypertension that should be recognized as an emergency and quickly treated. Although encephalopathy occurs mainly in patients with chronic or malignant hypertension, it can also complicate sudden hypertension of brief duration. The clinical manifestations are generated not only by the severity of BP elevation but also by the abrupt rise of BP in a previously normotensive individual. This condition occurs more frequently when the hypertension is complicated by renal insufficiency than when the renal function is normal. The full-blown clinical syndrome of hypertensive encephalopathy may take anywhere from 12 to 48 hours to develop.
Severe generalized, sudden headache is a prominent clinical manifestation. Neurogenic symptoms consisting of confusion, somnolence, and stupor may appear simultaneous with or following the onset of headache. If untreated, progressive worsening of neurological damage occurs, culminating in coma and death. The patient may be restless and uncooperative during the initial stages of the syndrome. Other clinical features may include projectile vomiting, visual disturbances ranging from blurring to frank blindness, and transient focal neurologic deficits. Sometimes (especially in children) generalized or focal seizures may be the only clinical feature.
On physical examination of the patient, the BP is markedly elevated but there is no certain level of BP above which encephalopathy is likely to develop. The fundi reveal generalized arteriolar spasm with exudates or hemorrhages. Although papilledema is present in most patients with this complication, its absence does not exclude the diagnosis of hypertensive encephalopathy.
When a patient with uncontrolled hypertension presents with severe headache, fluctuating mental status, papilledema, and variable neurologic deficits, the most likely initial diagnosis should be hypertensive encephalopathy. This must be distinguished from other acute neurologic complications of hypertension such as cerebral infarction or hemorrhage and uremic encephalopathy. A complete but quick evaluation of the patient should be carried out to consider various diagnostic possibilities. The only definitive clue to confirm the diagnosis of hypertensive encephalopathy is the prompt response of the patient's BP reduction. The syndrome should improve or reverse within a few hours with immediate control of the hypertension.
Once the diagnosis of hypertensive encephalopathy is likely, the BP should be lowered rapidly to near-normal levels; yet the diastolic BP should probably remain at or slightly above 100 mm Hg. Rapid reduction in the BP produces prompt, dramatic, and significant relief of the symptoms of hypertensive encephalopathy. The most important goal of therapy is to prevent permanent neurologic damage. Although potent orally effective agents like minoxidil and nifedipine can control severe hypertension, parenteral drugs are preferred in treating hypertensive encephalopathy successfully.1,6
Severe hypertension and stroke pose a difficult dilemma. When intracerebral pressure rises as a result of hemorrhage or thrombotic infarction, cerebral blood flow may no longer be under normal autoregulation. Therefore, a reduction in the systemic BP may conceivably compromise the cerebral flow. Conversely, persistence of severe hypertension may resolve spontaneously within 48 hours. There are no definitive data in the published literature that provide the practicing physician with a definitive approach or guidelines in managing these patients. At times it is difficult to make a distinction between an ischemic stroke and an intracerebral bleed.
With the current state of our knowledge, no reliable guidelines can be given for the management of hypertensive crises occurring in patients with cerebrovascular accidents. Based on the pathogenesis of these conditions, especially intracerebral hemorrhage, it is advisable to reduce the BP to near-normal levels or to a degree that will not clinically compromise cerebral function. If there is evidence of progression of the disease or worsening of the neurologic manifestations during treatment, then the therapeutic approach must be reassessed. Precautions should be taken to avoid hypotension in these patients, and it is advisable, therefore, not to lower the diastolic BP to less than 100 mm Hg. In any case, the BP reduction should be no more than the approximated 20% of the baseline BP level.
Acute aortic dissection has severe pain as its cardinal manifestation.7 This pain can easily be confused with the that of other conditions, such as acute MI. There are certain subtle differences, however: The pain of dissection is abrupt in onset, severe right from the beginning, and unremitting in most patients, whereas patients with acute MI rarely report abrupt onset of pain. The pain of MI may come and go, whereas the pain of aortic dissection is continuous. Once the diagnosis is suspected, immediate medical therapy should be implemented pending further diagnostic evaluation.
Two-dimensional echocardiography, transesophageal echocardiography, CT, tomography, angiography, and magnetic resonance angiography (MRA) have all been used to diagnose aortic dissection. MRA combined with transesophageal or transthoracic echocardiography yields considerable information. However, MRA is recommended only for patients who are hemodynamically stable. Digital and/or conventional angiography provides more thorough information concerning the anatomy of dissection.
Once the diagnosis of acute aortic dissection is apparent, the following steps should be undertaken. If the patient is hypertensive, the BP should be reduced to normal or below-normal levels with drugs that cause the BP to come down smoothly rather than drastically. Direct vasodilators that reflexively stimulate the heart should be avoided and they are contraindicated in acute aortic dissection. When considering medical therapy, keep in mind that the force and velocity of ventricular contraction (dP/dt) and the pulsatile flow are the determinants of the shearing force acting on the aortic wall. Attempts should be made to decrease the dP/dt with a suitable antihypertensive drug. Pharmacological options for acute aortic dissection include labetalol, a combined alpha- and beta-blocking drug, and the ultrashort-acting beta-blocker esmolol, in combination with nitroprusside.
Acute left ventricular failure may be caused by severe and uncontrolled hypertension; the higher the BP, the harder the left ventricle must work. Decreasing the workload of the failing myocardium may improve the heart function. In acute left ventricular failure, myocardial oxygen requirements increase due to increased end-diastolic fiber length and left ventricular volume. This could be particularly unfavorable in patients with concomitant coronary artery disease (CAD).
Prompt reduction of BP with a balanced vasodilating agent such as nitroprusside is indicated. Nitroprusside decreases both preload and afterload with restoration of myocardial function and cardiac output. Although the ACE inhibitors, by the virtue of their pharmacologic actions, may be useful in this situation, there is insufficient clinical experience concerning acute therapeutic response to these agents in patients with left ventricular failure. Morphine, diuretics, and other standard drugs are also used in this setting.
Severe hypertension associated with acute ischemic heart disease should benefit from BP reduction in theory, since systemic hypertension increases myocardial oxygen consumption by increasing the left ventricular wall tension. But there are no conclusive data to prove that acute treatment is beneficial. Reduction of systemic BP reduces the cardiac work, wall tension, and oxygen demand, and may thus limit myocardial necrosis in the early phase of infarction. With a reduction in the afterload, the hemodynamic status improves significantly in MI. Carefully supervised treatment of hypertension in patients with acute MI is, therefore, likely to be beneficial.
Pheochromocytoma hypertensive crisis may manifest with impressively dramatic clinical features. The BP is markedly elevated during the paroxysm and the patient may have profound sweating, marked tachycardia, pallor, numbness, tingling, and coldness of the feet and hands.8 A single attack may last from a few minutes to hours and may occur as often as several times a day to once a month, or less.
If pheochromocytoma crisis is suspected, the alpha-adrenergic blocking drug phentolamine (if available) should be given in a dose of 1 to 5 mg IV, to be repeated within a few minutes if needed. Nitroprusside would be an alternative to phentolamine, but phentolamine is more specific. A beta-blocking drug may be useful if the patient has a concomitant cardiac arrhythmia. Administration of beta-blocking agents should always be preceded by either phentolamine or pheno-oxybenzamine. Otherwise, beta blockade can aggravate the unopposed alpha-mediated peripheral vasoconstriction. Labetalol has also been advocated for this condition, but may not be generally effective in controlling the clinical manifestations of pheochromocytoma.
Clonidine withdrawal syndrome can result from abrupt discontinuation of a high-dosage regimen of clonidine, causing a hyperadrenergic state that mimics pheochromocytoma. When clonidine is abruptly discontinued (especially at high dosages) or rapidly tapered, a syndrome consisting of nausea, palpitation, anxiety, sweating, nervousness, and headache, along with marked elevation of the BP has been noted. Symptoms of clonidine withdrawal can be relieved by reinstituting the clonidine regimen. If there is marked elevation of BP and the patient is experiencing symptoms such as palpitations, chest discomfort, and epigastric discomfort, the IV administration of phentolamine or labetalol is recommended.
Monoamine oxidase inhibitors (MAOIs) can increase the risk of hypertensive crisis in patients who also take drugs such as ephedrine and amphetamines or consume foods containing large quantities of tyramine. In the presence of an inhibitor of monoamine oxidase, tyramine and indirectly acting sympathetic amines escape oxidative degradation, enter the systemic circulation, and potentiate the actions of catecholamines. Sympathomimetic amines such as those contained in nonprescription cold remedies can also provoke this response. Due to declining use of MAOIs, its reaction is rare.
Cocaine-induced hypertensive crisis can cause an abrupt, sudden increase in the systemic BP, resulting in a hypertensive emergency. Neurohumoral factors triggered by cocaine likely cause intense vasoconstriction and thus increase the vascular resistance and the BP. Sudden rise of BP in a previously normotensive individual may result in a serious cardiovascular complication. The BP should be lowered to safe limits without much delay.
Eclampsia is a potentially serious cardiovascular complication in a pregnant patient. Although the definitive therapy is delivery of the fetus, the BP should be reduced to prevent neurologic, cardiac, and renal damage. Although other antihypertensive drugs may be effective in reducing the BP, the agent of choice is hydralazine, which has a long record of safety. Animal studies have shown that nitroprusside can cause problems in the fetus; therefore, its use should be reserved for hypertension refractory to hydralazine or methyldopa. The ganglion-blocking drug trimethaphan should be avoided because of the risk of meconium ileus. In pregnancy-induced hypertension, volume depletion may be present and diuretics should be avoided. IV labetalol and hydralazine have been used to treat severe hypertension in pregnancy.9 ACE inhibitors and angiotensin-receptor blockers should be avoided due to possible fetal/placental toxicity. Magnesium sulfate is also used as adjunctive therapy to control the convulsions.
TREATMENT OF A PATIENT WITH HYPERTENSIVE EMERGENCY
Whether or not to hospitalize the patient, the therapeutic choices, and the choice of IV versus oral therapy depend on the clinical evaluation of the patient and available facilities. Patients with hypertensive emergencies should be hospitalized, and those with hypertensive urgencies may not always require admission to the hospital.
The therapeutic premise underlying the management of hypertensive emergency is not only to lower the BP quickly but to prevent, arrest, and reverse the target-organ damage. Therefore, close supervision of the patient is required while the BP is being lowered. There are no firm guidelines as to the degree of desired BP reduction, but a reasonable goal for most hypertensive emergencies is to lower the diastolic BP to 100 mm Hg (or to reduce the mean arterial pressure by 20%) over a period of minutes to hours.
Hypertensive emergencies should preferably be managed in an ICU to allow for continuous monitoring of the general hemodynamic status of the patient. Even though a secondary form of hypertension, such as renal artery stenosis or adrenal hypertension, may be a causative factor, the immediate goal should be to lower the BP to a safe level rather than undertaking a diagnostic workup. The rapidity of onset and duration of action of the chosen drug are important considerations in treating patients with hypertensive emergencies. The physician should be familiar with the hemodynamic and pharmacologic actions of the available drugs.
Role of concomitant diuretic therapy
Diuretics have a limited role in the management of hypertensive emergencies; however, they potentiate the therapeutic response of nondiuretic agents. When the BP does not respond satisfactorily to an adequate dose of the primary agent, adding a diuretic (such as furosemide) may be helpful. In volume overload states such as heart failure, concomitant administration of a loop diuretic is indicated for optimal results. Diuretics should not be used routinely in the management of hypertensive crises, as prior volume depletion may be present in some conditions, such as malignant hypertension. The need for diuretic therapy, therefore, should be individualized on the basis of the hemodynamic and renal function status of the patient.
Orally effective drugs
Clinical experience has shown that antihypertensive drugs given orally as either single or multiple doses can lower the BP immediately in patients with severe hypertension. Obviously, this therapeutic opportunity is most suitable for patients with hypertensive urgencies, not emergencies.
Nifedipine, a calcium channel blocker given either orally or sublingually, has been shown to lower the BP rapidly and can be useful in the management of hypertensive crisis. Immediate reduction in BP can be accomplished with sublingual (a punctured capsule, or nifedipine liquid drawn out of the capsule with a syringe) or oral administration of the capsules. The drug is also effective when the capsule is bitten and then swallowed. The advantages of nifedipine are rapid onset of action and lack of CNS depression. It may cause reflex tachycardia. As the duration of action of nifedipine is short, patients who receive this drug for hypertensive emergencies should be monitored for several hours to consider re-administration of the drug. An abrupt fall in the BP induced by nifedipine administration can cause certain adverse effects—symptomatic hypotension, tachycardia, and ischemic events—so the clinical need to use nifedipine capsules to lower the BP urgently should be carefully assessed and avoided, if possible.
Clonidine has been shown to produce an immediate antihypertensive effect when given in repetitive doses. Typically, clonidine loading was accomplished in the ED by administering it orally at a dosage of 0.1 mg q1h until the desired goal was obtained. The use of clonidine loading therapy has declined due to the availability of safer and better-tolerated alternative therapies that do not cause drowsiness.
Captopril is an ACE inhibitor that has been found to be effective in the immediate treatment of severe hypertension and hypertensive crises. Captopril lowers the BP promptly without causing tachycardia, and thus offers a distinct hemodynamic advantage over direct arteriolar dilators. The maximal effect from orally administered captopril may not be attained for as long as 2 hours. There are some reports documenting the effectiveness of sublingual captopril in the treatment of hypertensive crisis, but further data needs to be generated to define its role in the acute management of hypertensive crisis.
Minoxidil, a powerful direct vasodilator, has been successfully used in the treatment of refractory or severe hypertension. Because of its relatively rapid onset of action and sustained duration, this drug has been used for the treatment of hypertensive crises. Minoxidil in dosages ranging from 2.5 to 10 mg can be given every 4 to 6 hours initially in the treatment of severe hypertension. It works best when given along with a diuretic, and an adrenergic blocker is necessary to counteract the reflex tachycardia. Minoxidil should be used with caution in patients with acute coronary syndromes unless the patient is beta-blocked.
Labetalol can be administered orally in doses of 100 to 300 mg for the treatment of hypertensive urgencies. Because of its dual adrenergic blockade, the fall in BP is not accompanied by reflex tachycardia. This attribute can be especially beneficial in patients with CAD. Oral labetalol is effective within 1 to 3 hours and may be useful for hypertensive urgencies, but it has an unpredictable dose response and thus may not be ideal for acute situations. Labetalol is contraindicated in patients with bronchospasm, heart block, significant bradycardia or heart failure.
IV drugs for hypertensive emergencies Several IV drugs are effective for use in hypertensive crisis. The patient's clinical presentation will determine the agent of choice in an individual situation. Parenteral drugs for rapid control of severe hypertension are shown in Table 3.
Nitroprusside is a powerful BP-lowering drug that possesses the properties of rapid onset and offset of action. The hypotensive response occurs within seconds after the infusion is started and disappears almost as rapidly when the infusion is discontinued. The initial infusion rate should be 0.3 mcg/kg/min, which can be increased every 5 minutes until a desired BP level is obtained. Once the desired effect is achieved, the BP should be continuously monitored. Hypotension is the most common—but avoidable—adverse effect of nitroprusside therapy. Cyanide toxicity from nitroprusside is possible, although extremely rare. Prophylactic infusion of cyanocobalamin (vitamin B12a) 25 mg/h, has been shown to decrease the cyanide concentration and tissue hypoxia resulting from nitroprusside infusion during surgery. Thiocyanate toxicity secondary to nitroprusside is uncommon and occurs only with high doses and in the presence of renal failure. Treatment should be interrupted when the thiocyanate level is close to 10 mg/dL. Monitoring of plasma thiocyanate levels is not mandatory, as long as the patient's clinical status is closely assessed. Treatment of suspected thiocyanate toxicity requires discontinuation of the drug and institution of dialysis.
Labetalol can be used parenterally or orally for the treatment of hypertensive emergencies. IV labetalol administered as either a continuous infusion or as bolus injections reduces the BP promptly because of its rapid onset of action. Controlled smooth reduction in BP may be obtained by continuous infusion of labetalol at the rate of 0.5 to 2 mg/min. As with nitroprusside, close monitoring of the patient is required during the infusion of labetalol. Rapid (but not abrupt) lowering of BP can also be accomplished with bolus injections of labetalol. Labetalol should not be used when beta-blockers are contraindicated—for example, in patients who may have heart failure, atrioventricular block, asthma, or chronic obstructive pulmonary disease.
Nicardipine is a dihydropyridine calcium antagonist that exerts a prompt hemodynamic effect when given IV in patients with severe hypertension. Nicardipine infusion is started at 5 mg/h, and can be titrated upwards gradually to obtain the desired therapeutic effect.11 Once a stable BP is reached, most patients do not require further dosage alterations. Thus, nicardipine pharmacodynamics resemble those of nitroprusside in terms of the onset, and duration and offset of action. Because of its mechanism of action (calcium channel blockade), nicardipine may be beneficial in preserving tissue perfusion. It is a good option in the management of severe hypertension with or without target-organ damage.
Hydralazine has an antihypertensive action that results from a direct relaxation of the vascular smooth muscle and it is accompanied by reflex increases in stroke volume and heart rate, which can precipitate myocardial ischemia. Either IM or IV administration of hydralazine causes an unpredictable but definite fall in BP. In the treatment of hypertensive emergencies, the initial IV dose should be 10 to 20 mg. The onset of the hypotensive effect occurs within 10 to 30 minutes, and its duration of action ranges from 3 to 9 hours. The dose and frequency of administration necessary to control the BP are highly variable. The delayed onset and unpredictable degree of hypotensive effect present difficulties in titration. Nevertheless, hydralazine continues to be successfully utilized in the treatment of eclampsia.
Phentolamine is an alpha receptor-blocking agent that is specifically indicated for treating hypertensive crises associated with increased circulating catecholamines. These include, for example, pheochromocytoma crisis, certain cases of clonidine withdrawal syndrome, and crises resulting from an MAOI and drug-food interaction. The hypotensive effect of a single bolus injection is short-lived and lasts less than 15 minutes. This drug may precipitate angina or cardiac arrhythmias.
Nitroglycerin is a weak systemic arterial dilator with a greater effect on large arteries than on smaller arteries. Low doses cause venodilation; much higher doses are required to produce a fall in systemic BP. Because of its pharmacologic actions, nitroglycerin infusion may be particularly beneficial in patients with CAD either with or without hypertension. Nitroglycerin improves coronary perfusion. Although there are no controlled studies, nitroglycerin therapy can be considered in the management of severe hypertension associated with CAD. The usual initial dosage of nitroglycerin is 5 to 15 mcg/min, and the dosage is titrated upward to a desired therapeutic end point. It has a rapid onset (2-5 minutes) and offset of action. Although nitroglycerin has been used to achieve controlled hypotension, its main use continues to be in patients with unstable angina and acute MI. Prolonged use may result in tolerance. Isosorbide dinitrate therapy has also been utilized for immediate treatment of severe hypertension but its precise role and guidelines on how to use it are not fully delineated.12
Enalaprilat, by its mechanism of action, prevents conversion of angiotensin I to angiotensin II by blocking angiotensin converting enzyme and thus lowering BP. Enalaprilat is the only available parenteral ACE inhibitor. For hypertensive emergencies, it is given at a dosage of 0.625 to 1.25 mg IV over 5 minutes and may be repeated every 6 hours. ACE inhibitors are contraindicated in patients with renal artery stenosis and in pregnant patients. ACE inhibitors can cause precipitous falls in BP in patients who are hypovolemic. These drugs are especially valuable in treating hypertensive emergencies in patients with chronic heart failure.
Fenoldopam is a relatively new drug that is a selective dopamine (DA1) receptor agonist and causes significant vasodilation. It is a short-acting, parenteral arteriolar vasodilator that lowers BP by reducing peripheral vascular resistance.13,14 By the mode of its action on the DA1 receptor, fenoldopam causes remarkable renal vasodilation and promotes diuresis and natriuresis. The half-life of fenoldopam is estimated to be 5 minutes, and the recommended starting dosage is 0.1 mcg/ kg/min. Its onset of action is 5 minutes and maximum response is achieved within 15 minutes. The dosage should be titrated to gradually attain the goal BP. It can lower BP in hypertensive emergencies safely and effectively and at the same time preserve or improve renal blood flow and cause diuresis and natriuresis. Therefore, it may offer important advantages. A number of early comparative studies between nitroprusside and fenoldopam showed that while both are equally effective in reducing the BP, fenoldopam offers important renal advantages over nitroprusside.
Affecting one quarter of the adult population (60 million in the United States and 1 billion people worldwide), arterial hypertension is the leading cause of death in the world and the most common cause for an outpatient visit to a physician; it is the most easily recognized treatable risk factor for stroke, myocardial infarction, heart failure, peripheral vascular disease, aortic dissection, atrial fibrillation, and end-stage kidney disease. Despite this knowledge and unequivocal scientific proof that treatment of hypertension can prevent many of its life-altering complications, hypertension remains untreated or undertreated in the majority of affected individuals in all countries, including those with the most advanced systems of medical care. Inadequate treatment of hypertension is a major factor contributing to some of the adverse secular trends since the early 1990s, including an increased incidence of stroke, heart failure, and kidney failure plus a leveling off of the decline in coronary heart disease mortality.
Across populations, the risks of heart disease and stroke increase continuously and logarithmically with increasing levels of systolic and diastolic blood pressure at or above 115/75 mm Hg . Thus, the dichotomous separation of “normal” from “high” blood pressure is artificial, and the definition of arterial hypertension (i.e., high blood pressure) has been a moving target. On the basis of results of randomized clinical drug trials, hypertension currently is defined as a usual blood pressure of 140/90 mm Hg or higher, the value above which the benefits of treatment appear to outweigh the risks. Prehypertension is a new designation for mildly elevated blood pressures between 120/80 and 139/89 mm Hg, a level at which progression to hypertension is twice as likely as with a blood pressure below 120/80 mm Hg, and cardiovascular risk retains its continuous log-linear function compared with lower blood pressures. The cardiovascular mortality rate is only half as great at 120/80 mm Hg as at 140/90 mm Hg, but it is unknown whether the benefits of treating prehypertension outweigh the risks.
The asymptomatic nature of the condition impedes early detection, which requires regular blood pressure measurement. Because most cases of hypertension cannot be cured, blood pressure control requires lifelong treatment with prescription medication, which is costly and often causes more symptoms than the underlying disease process. Effective management requires continuity of care by a regular and knowledgeable physician as well as sustained active involvement by an educated patient.
Arterial hypertension, defined as a systolic blood pressure (SBP) in excess of 140 mm Hg and/or diastolic blood pressure (DBP) in excess of 90 mm Hg, has long been identified as an independent risk factor for cardiovascular disease. Traditionally, emphasis has been placed on elevated DBP as a risk factor for the development of target organ damage. However, as early as 1971, the Framingham study showed that, although DBP was a major determinant of cardiovascular risk in men under 45 years of age, SBP was the stronger risk factor in older men and in women of all ages. Since then, several observational studies have suggested that the pulse pressure (PP) may be a better predictor of cardiovascular complications than SBP or mean arterial pressure.
Data from the National Health and Nutrition Examination Survey has demonstrated that if a blood pressure (BP) of 140/90 mm Hg is considered to be normal, only 27% of hypertensive patients are adequately controlled in the United States. Recommendations from the Joint National Committee on the Prevention, Detection, Evaluation and treatment of High Blood Pressure (JNC-VI report) now regard a BP of 140/90 mm Hg as high normal and 130/85 mm Hg as normal. For diabetic patients, therefore, it is recommended that BP be reduced below 130/85 mm Hg and for those with renal impairment, evidenced by proteinuria, pressures should be reduced below 125/75 mm Hg. In patients with underlying coronary artery disease, the BP should be reduced below 120/80 mm Hg.
The beating heart generates pressure and flow waves which propagate throughout the arterial system. The shape of the pressure and flow waves is altered by their continuous interaction with the non-uniform arterial system. The pressure and flow waves can be studied in terms of a forward component, running from the heart itself, and a backward component carrying information on the peripheral arterial system.
In the presence of arteriosclerosis and aortic stiffening (consequences of arterial hypertension), the pulse wave velocity is increased, causing the pulse waves to reflect more quickly off the arteriolar vessels and return to the large vessels during systole. This amplifies SBP. In the presence of normal vascular compliance, the reflected waves return during diastole and augment DBP. Consequently, arteriosclerosis tends simultaneously to increase SBP and decrease DBP, resulting in a widened pulse pressure.
A widened pulse pressure increases cardiovascular morbidity because elevated SBP is associated with greater left ventricular workload and myocardial oxygen demand, whereas a decreased DBP may decrease coronary perfusion, resulting in decreased myocardial oxygen supply and a greater risk for myocardial ischemia and injury.
Hypertension (systolic pressure 140 mm Hg or diastolic pressure 90 mm Hg) is present in one in four adults in the United States.1The prevalence is higher among blacks and older persons, especially older women. Table 1 shows the classification of blood pressure according to the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure.2Hypertension is a risk factor for stroke, myocardial infarction, renal failure, congestive heart failure, progressive atherosclerosis, and dementia.3 Systolic pressure is a stronger predictor of cardiovascular events than is diastolic pressure,4 and isolated systolichypertension, which is common among older persons, is particularly hazardous.5 There is a continuous, graded relation between blood pressure and the risk of cardiovascular disease; the level and duration of hypertension and the presence or absence ofcoexisting cardiovascular risk factors determine the outcome.6 Treatment of hypertension reduces the risk of stroke, coronary artery disease, and congestive heart failure, as well as overall cardiovascular morbidity and mortality from cardiovascular causes. However, only 54 percent of patients with hypertension receive treatment and only 28 percent have adequately controlled blood pressure.1
Strategies and Evidence
Accurate measurement of blood pressure7 and verification of elevated pressure on multiple occasions over time are important.Ambulatory or home blood-pressure monitoring8 can identify "white-coat hypertension" (blood pressure that is elevated when measured during an office visit but that is otherwise normal) and prevent unnecessary treatment. White-coat hypertension, present in 20 percent of patients with elevated blood pressure, is associated with a lower cardiovascular risk than is sustained hypertension, but it may be a precursor of sustained hypertension and therefore warrants monitoring.
In addition to the history taking and physical examination, several tests are routinely indicated in patients with hypertension: urinalysis, complete blood count, blood chemical tests (measurements of potassium, sodium, creatinine, fasting glucose, total cholesterol, and high-density lipoprotein), and 12-lead electrocardiography. The evaluation should identify signs of cardiovascular, cerebrovascular,or peripheral vascular disease and other cardiovascular risk factors that are frequently present in patients with hypertension. Severe or resistant hypertension or clinical or laboratory findings suggesting the presence of renal disease, adrenal hypertension (due to abnormal mineralocorticoid secretion or pheochromocytoma), or renovascular hypertension should be further investigated. Essential, or primary, hypertension, the focus of this article, is the diagnosis in over 90 percent of cases.
The primary goal of the treatment of hypertension is to prevent cardiovascular disease and death. Coexisting cardiovascular risk factors increase the risks associated with hypertension and warrant more aggressive treatment. The five-year risk of a major cardiovascular event in a 50-year-old man with a blood pressure of 160/110 mm Hg is 2.5 to 5.0 percent; the risk doubles if the man has a high cholesterol level and triples if he is also a smoker.9
The benefits of lowering blood pressure, first demonstrated after short-term treatment of malignant hypertension,10 havesubsequently been demonstrated in all stages of hypertension. Trials involving patients with stage 1 or 2 hypertension showed that lowering systolic pressure by 10 to 12 mm Hg and diastolic pressure by 5 to 6 mm Hg reduces the risk of stroke by 40 percent, the risk of coronary disease by 16 percent, and the risk of death from any cardiovascular cause by 20 percent.11,12 The higher the blood pressure and the number of risk factors, the greater the reduction in absolute risk (and the smaller the number needed to treat).
Determination of the need for drug therapy is based on a combined assessment of the blood-pressure level and the absolute risk ofcardiovascular disease (Figure 1). Patients with stage 1 hypertension can be treated with lifestyle modifications alone for up to one year, if they have no other risk factors, or for up to six months, if they have other risk factors. Drug treatment should be provided if blood pressure remains elevated after a trial of lifestyle modifications alone. Lifestyle modifications and antihypertensive therapy are indicated for patients with cardiovascular or other target-organ disease (renal, cardiac, cerebrovascular, or retinal disease) and for those with stage 2 or 3 hypertension. Patients with diabetes are at high risk, and drug therapy is indicated in such patients even if blood pressure is at the high end of the normal range.
Figure 1. Treatment of Hypertension According to the Level of Blood Pressure and Cardiovascular Risk.
Two or more blood-pressure readings separated by two minutes should be averaged. If the pressure is at the high end of the normal range, it should be rechecked yearly. Stage 1 hypertension should be confirmed within two months. Patients with stage 2hypertension should be evaluated and referred for care within one month; those with stage 3 hypertension should be evaluated immediately or within one week. If systolic and diastolic values are in different categories, the recommendations for the higher reading should be followed.
Laboratory tests include a complete blood count; measurements of potassium, sodium, creatinine, fasting glucose, total cholesterol, and high-density lipoprotein cholesterol; and urinalysis. ECG denotes electrocardiography. Cardiovascular or other target-organ disease denotes left ventricular hypertrophy, angina or prior myocardial infarction, prior coronary revascularization, heart failure, stroke or transient ischemic attack, nephropathy, peripheral arterial disease, and retinopathy.
For patients with multiple risk factors, clinicians should consider drugs as initial therapy along with lifestyle modifications. Clinically important risk factors include smoking, dyslipidemia, diabetes mellitus, an age of more than 60 years, male sex, postmenopausal status in women, and a family history of cardiovascular disease for women under the age of 65 years and men under the age of 55 years. Adapted from the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure.2
Table 2 lists lifestyle modifications recommended for all patients with hypertension. The Dietary Approaches to Stop Hypertension(DASH) study showed that eight weeks of a diet of fruits, vegetables, low-fat dairy products, whole grains, poultry, fish, and nuts,with limited fats, red meat, and sweets, reduced systolic pressure by 11.4 mm Hg and diastolic pressure by 5.5 mm Hg.13 With sodium intake at a level below 100 mmol per day, systolic pressure was 3 mm Hg lower and diastolic pressure was 1.6 mm Hg lowerthan with the DASH diet and a higher level of sodium intake.14
Restriction of sodium intake to 2 g per day lowers systolic pressure, on average, by 3.7 to 4.8 mm Hg and lowers diastolic pressure, on average, by 0.9 to 2.5 mm Hg,15,16 although the reductions vary from person to person beyond these ranges. Salt sensitivity is common in elderly patients with hypertension. Despite concern that salt restriction for all patients with hypertension might have adverse consequences,17 moderate sodium restriction appears to be generally safe and effective18 and is particularly effective in elderly persons.19
Whether lifestyle modifications can be sustained is a concern. Four years after enrollment in the Treatment of Mild HypertensionStudy, patients with stage 1 hypertension had gained back half the weight lost after one year of intervention and were less successful at maintaining a low sodium intake and an increased level of physical activity than they had been at one year.20 Nevertheless, lifestyle modifications alone controlled blood pressure at four years in 59 percent of the patients.
Most clinical trials of lifestyle modifications have been underpowered or of insufficient duration to evaluate the effect of theseinterventions on major cardiovascular outcomes. However, lifestyle modifications should be encouraged, since they are safe andinexpensive and, when combined with drug therapy, may result in better blood pressure control and an improved quality of life.21
Treatment Goal for Blood Pressure
The risk of cardiovascular disease remains higher in treated patients with hypertension than in persons with normal blood pressure, suggesting that treatment targets have not been low enough. Greater reductions in blood pressure have been shown to be safe and beneficial.22,23 In the Hypertension Optimal Treatment trial, the risk of major cardiovascular events was lowest among patients whose blood pressure had been reduced to 138.5/82.6 mm Hg. An additional reduction did not further reduce the risk of events in nondiabetic patients, but it was not harmful. Among diabetic patients, the lowest rates of major cardiovascular events and death from cardiovascular causes were achieved with the lowest blood pressure. In patients over the age of 65 years, morbidity and mortality from cardiovascular disease are reduced when systolic pressure is lowered to a level below 160 mm Hg.24 Whether levels below 140 mm Hg provide additional protection is unclear.
Choice of Antihypertensive Drugs
Most antihypertensive drugs reduce blood pressure by 10 to 15 percent. Monotherapy is effective in about 50 percent of unselectedpatients, and those with stage 2 or 3 hypertension often need more than one drug.25 There have been few comparative trials ofantihypertensive agents that have had sufficient power to demonstrate an advantage of one drug over another, and there is individual variation in responsiveness to drugs. Thus, the choice of therapy is based on a combined assessment of several characteristics of the patient: coexisting conditions, age, race or ethnic group, and the response to previously used drugs, including the presence or absence of adverse reactions.
A critical issue is whether a drug reduces cardiovascular morbidity and mortality. As compared with placebo, diuretics and beta-blockers reduce the risk of stroke, coronary heart disease, and overall mortality from cardiovascular disease in unselected patientswith hypertension who do not have preexisting coronary disease, diabetes, or proteinuria.11,12 A meta-analysis of trials involvingmore than 26,000 patients showed that, as compared with placebo, angiotensin-converting–enzyme (ACE) inhibitors reduce the riskof stroke, coronary heart disease, major cardiovascular events, death from cardiovascular causes, and death from any cause,26although the results were heavily dependent on a trial in which all the participants had preexisting cardiovascular disease or diabetes and some did not have hypertension.27 Calcium-channel antagonists, as compared with placebo, reduce the risk of stroke, major cardiovascular events, and death from cardiovascular causes; however, these drugs do not significantly reduce the risk of coronary heart disease, heart failure, or death from any cause.26
The question of whether antihypertensive agents differ in their ability to prevent adverse outcomes has been difficult to answer.28Some data suggest potentially important differences. For example, ACE inhibitors were more effective than calcium-channel antagonists in preventing coronary heart disease in one trial,29 but not in another, larger study.30 A meta-analysis of clinical trialssuggests that ACE inhibitors are more effective than calcium-channel antagonists in reducing the risk of heart failure but not inreducing the risk of stroke, death from cardiovascular disease, or death from any cause.26 Losartan, an angiotensin-receptorantagonist, has recently been shown to be more effective than atenolol in reducing the risk of stroke.31 Another meta-analysissuggests that calcium-channel antagonists may prevent stroke to a greater extent than diuretics or beta-blockers but have not been shown to provide similar protection against coronary heart disease.32 The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial, the largest randomized trial comparing several antihypertensive agents as initial therapy, demonstrated that in patients older than 55 years (35 percent of whom were black and 19 percent of whom were Hispanic), diuretic-based therapy was as effective as treatment with calcium-channel antagonists or ACE inhibitors in preventing major coronary events.33 Diuretic-basedtherapy was slightly more effective than treatment with calcium-channel antagonists in preventing heart failure and was more effectivethan treatment with ACE inhibitors in preventing stroke and heart failure. A smaller study of elderly white men and women withhypertension, showed that ACE-inhibitor–based therapy was slightly more effective than diuretic-based therapy in preventing myocardial infarction (only in men) but not stroke.34
On the basis of the available data, diuretics or beta-blockers remain appropriate for the initial treatment of uncomplicatedhypertension, despite the concern that these agents may be associated with adverse metabolic effects (e.g., hyperuricemia and impaired glucose tolerance). Alternative drugs are preferable for patients with certain coexisting medical conditions (Table 3). In particular, ACE inhibitors and angiotensin-receptor antagonists are appropriate initial therapy in patients with diabetes mellitus, renal disease, or congestive heart failure35,36 (though beta-blockers and diuretics are also useful in patients with heart failure); ACE inhibitors can also be used in patients with prior myocardial infarction or coronary artery disease. Short-acting calcium-channel antagonists cause a rapid, acute drop in blood pressure, which may precipitate coronary ischemia, and long-acting calcium-channel antagonists are therefore preferred when this class of agent is chosen.37 Alpha-blockers relieve symptoms associated with prostatic hypertrophy. Since they are not as effective as other agents in reducing the risk of cardiovascular disease, they should be used as second- or third-line therapy.33
Other Considerations in the Choice of Therapy
Age and race have been shown to be determinants of the response to specific antihypertensive medications. The Department ofVeterans Affairs Cooperative Study reported that younger whites had a good response to ACE inhibitors and beta-blockers, whereasolder blacks had a better response to diuretics or calcium-channel antagonists.25
Hypertension is more severe and target-organ damage, particularly end-stage renal disease, more prevalent among blacks. Salt sensitivity is common, and sodium restriction should be encouraged. Although the magnitude of the blood-pressure response to monotherapy with a diuretic or a calcium-channel antagonist may be greater than the response to monotherapy with another agent, significant reductions occur with ACE inhibitors, angiotensin-receptor antagonists, and beta-blockers when an adequate dose is given.38
Side effects differ according to the class of antihypertensive drug (Table 3). Although adverse effects are reported by 10 to 20 percent of patients taking such drugs, the quality of life improves when hypertension is treated.21 The Treatment of MildHypertension Study and the Department of Veterans Affairs Cooperative Study both demonstrated that among the five main classesof antihypertensive drugs (diuretics, beta-blockers, calcium-channel antagonists, ACE inhibitors, and alpha-blockers), no one drug is more acceptable than the others, except that sexual dysfunction is more common among men treated with the diuretic chlorthalidone.21,25 Use of lower-cost, generic drugs that require less frequent doses can improve compliance.
The use of lower doses of two or more drugs with complementary mechanisms may lower blood pressure with fewer adverse effectsthan the use of higher doses of a single agent. Most combination therapies include small doses of a diuretic, which potentiate the effects of other drugs (ACE inhibitors, angiotensin-receptor antagonists, or beta-blockers). Combination therapy may improvecompliance and achieve the target blood pressure more rapidly.
National and international groups have issued guidelines for the treatment of hypertension. The main differences among theseguidelines are the criteria for initiating drug therapy in low-risk patients with stage 1 hypertension. The Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure2 and the World Health Organization–International Societyof Hypertension41 recommend stratification of patients into risk categories on the basis of age, sex, smoking status, presence or absence of diabetes, cholesterol level, presence or absence of preexisting cardiovascular disease, and presence or absence of target-organ damage (Figure 1). Drug treatment is recommended for stage 1 or higher hypertension if blood pressure does not decrease after a certain period of lifestyle-modification counseling (6 to 12 months, according to the Joint National Committee guidelines). The British Hypertension Society and New Zealand guidelines recommend the use of tables that quantify a person's 5- or 10-year risk of a cardiovascular event; drugs are recommended only if the 5-year risk is at least 10 percent.42,43 When drugs are indicated, the guidelines recommend those that have been shown to improve cardiovascular outcomes, with coexisting conditions and demographic characteristics taken into account.
Areas of Uncertainty
Although moderate sodium restriction lowers blood pressure, the small effects, variability in response, and lack of a provencardiovascular benefit have led to uncertainty about whether it should be broadly recommended. There is also uncertainty about whether specific properties of certain drugs result in differential effects on morbidity and mortality that are independent of the reduction in blood pressure.
The use of drugs in patients with a low absolute risk of cardiovascular disease is controversial. The rationale for withholding drugsfrom such patients is that some trials have shown that mortality among low-risk patients treated with drugs is similar to that in control groups.44 However, given that even high-normal blood pressures (130 to 139/85 to 89 mm Hg) are associated with an increased riskof cardiovascular disease,45 there is concern about withholding drugs from "low-risk" patients. Also, the feasibility of basingtreatment decisions on the use of tables for calculating the absolute risk of cardiovascular disease has not been assessed.
The appropriate strategy for choosing the initial antihypertensive therapy is still unresolved. Some have proposed that the choice oftreatment should be based on renin levels,46 but this approach is not widely used. Whether combination therapy as the initialtreatment leads to better control of blood pressure and a lower risk of cardiovascular disease than monotherapy is also unresolved.Finally, optimal blood-pressure targets remain to be determined, particularly for elderly patients.
Conclusions and Recommendations
Hypertension affects 25 percent of adults in the United States and is adequately treated in less than 30 percent of them. Appropriatetherapy can reduce blood pressure and cardiovascular morbidity and mortality.
Persons who have stage 1 hypertension and are at low risk for cardiovascular disease can be treated with lifestyle modifications for up to one year. Patients who have stage 1 hypertension and other cardiovascular risk factors or a higher stage of hypertension should be treated with drugs to reduce blood pressure to a level below 140/90 mm Hg, or to reduce pressure to 130/80 mm Hg or less if the patient has diabetes, renal disease, or both.
Diuretics and beta-blockers are appropriate as first-line therapy for patients without coexisting conditions. ACE inhibitors orangiotensin-receptor antagonists are recommended for patients with type 2 diabetes, kidney disease, or both and are also useful in patients with heart failure. Beta-blockers and ACE inhibitors are recommended in patients with prior myocardial infarction, and calcium-channel antagonists benefit elderly patients at risk for stroke. If blood pressure is not controlled with an optimal dose of a single drug, a second agent with a complementary mechanism of action should be added. Combination therapy provides more rapid control of blood pressure than does monotherapy and is therefore an initial treatment option for patients with stage 2 or 3hypertension.
The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressureemphasizes the need to consider secondary causes, improper blood-pressure measurement, volume overload, competing substances, obesity, nonadherence to treatment, inadequate doses or inappropriate combinations of medications, and alcohol consumption as factors in resistant hypertension.1 These guidelines do not include specific recommendations regarding when or how to evaluate patients for specific secondary causes of resistant hypertension or for the management of truly resistant cases.
Areas of Uncertainty
Several questions require further investigation.37 The true prevalence of resistant hypertension remains uncertain. More information is needed to determine the optimal evaluation of patients for secondary hypertension, including indications for screening for hyperaldosteronism, which appears to be underdiagnosed. Data from randomized trials are needed to improve the treatment of patients whose blood pressure remains high while they are receiving multiple agents. In such patients, the possible role of new drugs, such as inhibitors of renin and endothelin-1, requires evaluation.
Summary and Recommendations
The patient in the vignette has elevated blood pressure, despite taking three medications, with evidence of target-organ injury (retinal arteriopathy and left ventricular hypertrophy). Careful assessment of her adherence to therapy is warranted. Such adherence may be improved by addressing the reasons for nonadherence, such as side effects or the cost of medications, or by arranging for more frequent office visits or telephone contact. The ibuprofen she takes for osteoarthritis should probably be discontinued, since it may contribute to blood-pressure elevation, and be replaced with acetaminophen. Weight loss and a restriction of dietary sodium should be encouraged. Since volume overload is common in refractory hypertension, she could increase her dose of diuretic (with repletion of potassium as needed). If these interventions are ineffective, a different class of drug (e.g., a calcium-channel blocker) could be added, and the patient could be screened for renovascular hypertension, even though in patients with this condition, blood pressure may remain elevated despite intervention. Regular follow-up is warranted, with a goal of maintaining the blood pressure below 140/90 mm Hg.
Hypertensive retinopathy is a condition characterized by a spectrum of retinal vascular signs in people with elevated blood pressure.1The detection of hypertensive retinopathy with the use of an ophthalmoscope has long been regarded as part of the standardevaluation of persons with hypertension.2,3,4 This clinical practice is supported by both previous5 and current6 reports of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC), which list retinopathy as one of several markers of target-organ damage in hypertension. On the basis of the JNC criteria, the presence of retinopathy may be an indication for initiating antihypertensive treatment, even in people with stage 1 hypertension (blood pressure, 140 to 159/90 to 99 mmHg) who have no other evidence of target-organ damage.
Despite the JNC recommendation, the clinical implications of hypertensive retinopathy are unclear. Many physicians do not regularly perform an ophthalmoscopic examination as part of the care they provide to hypertensive patients, nor do they include retinal findings when making decisions about treatment. Furthermore, there is no clear consensus regarding the classification of hypertensiveretinopathy or whether a retinal examination is useful for risk stratification.
The evidence in support of the JNC guidelines on retinal findings in hypertension is based on earlier studies that may not have direct relevance to current clinical practice.7,8,9,10 These studies have several important limitations. First, because they involved patients who had uncontrolled and untreated hypertension, generalization to contemporary populations of patients with lower blood-pressure levels may be problematic. Second, retinopathy as defined in these studies was based on a direct ophthalmoscopic examination. This technique has been shown to be unreliable, with high rates of interobserver variability (20 to 42 percent) and intraobserver variability (10 to 33 percent) when used in persons with mild hypertension.11,12 Third, although many earlier studies cite increased mortality among persons with hypertensive retinopathy,8,9,10 few studies have demonstrated associations between hypertensive retinopathyand specific cardiovascular outcomes (e.g., incident stroke and coronary heart disease) or have adequately controlled for relevant confounding factors (e.g., hyperlipidemia and cigarette smoking). Thus, whether hypertensive retinopathy predicts the risk of cardiovascular outcomes independently of other risk indicators has not been examined until recently. The purpose of this review is to appraise recent studies (i.e., from 1990 onward) in regard to the pathophysiology, epidemiology, and cardiovascular associations ofhypertensive retinopathy and the evidence that supports its use for risk stratification in persons with hypertension.
Historical Context and Classification
Hypertensive retinopathy was first described by Marcus Gunn in the 19th century in a series of patients with hypertension and renal disease.7 The retinal signs he observed included generalized and focal arteriolar narrowing, arteriovenous nicking, flame-shaped and blot-shaped retinal hemorrhages, cotton-wool spots, and swelling of the optic disk (Figure 1, Figure 2, and Figure 3). In 1939, Keith et al. showed that these signs of retinopathy were predictive of death in patients with hypertension.10 The authors described a widely used classification system that categorized these signs into four groups of increasing severity.
Examples of Mild Hypertensive Retinopathy.Panel A shows arteriovenous nicking (black arrow) and focal narrowing (white arrow). Panel B shows arteriovenous nicking (black arrows) and widening or accentuation ("copper wiring") of the central light reflex of the arterioles (white arrows).
Examples of Moderate Hypertensive Retinopathy. Panel A shows retinal hemorrhages (black arrows) and a cotton-wool spot (white arrow). Panel B shows cotton-wool spots (white arrows) and arteriovenous nicking (black arrows).
However, several reviews of hypertensive retinopathy since 1996 have questioned the usefulness of the classification system by Keith et al. (subsequently modified by Scheie) and its relevance to current clinical practice. The major criticisms of the original and modified classifications are that they do not enable the clinician to distinguish among low retinopathy grades (e.g., grade 1 signs are not easily distinguished from grade 2 signs) and that the retinopathy grades are not closely correlated with the severity of hypertension. Furthermore, a detailed categorization of retinopathy into four grades does not appear to be supported by retinal studies with the use of fluorescein angiography.
The retinal circulation undergoes a series of pathophysiological changes in response to elevated blood pressure. In the initial,vasoconstrictive stage, there is vasospasm and an increase in retinal arteriolar tone owing to local autoregulatory mechanisms. This stage is seen clinically as a generalized narrowing of the retinal arterioles. Persistently elevated blood pressure leads to intimal thickening, hyperplasia of the media wall, and hyaline degeneration in the subsequent, sclerotic, stage. This stage corresponds to more severe generalized and focal areas of arteriolar narrowing, changes in the arteriolar and venular junctions (i.e., arteriovenous nicking or nipping), and alterations in the arteriolar light reflex (i.e., widening and accentuation of the central light reflex, or "copper wiring").
This is followed by an exudative stage, in which there is disruption of the blood–retina barrier, necrosis of the smooth muscles and endothelial cells, exudation of blood and lipids, and retinal ischemia. These changes are manifested in the retina as microaneurysms,hemorrhages, hard exudates, and cotton-wool spots. Swelling of the optic disk may occur at this time and usually indicates severely elevated blood pressure (i.e., malignant hypertension). Because better methods for the control of blood pressure are now available in the general population, malignant hypertension is rarely seen. In contrast, other retinal vascular complications of hypertension, such as macroaneurysms and branch-vein occlusions, are not uncommon in patients with chronically elevated blood pressure. The stages ofhypertensive retinopathy described here, however, may not be sequential. For example, signs of retinopathy that reflect the exudative stage, such as retinal hemorrhage or microaneurysm, may be seen in eyes that do not have features of the sclerotic stage (e.g., arteriovenous nicking). The exudative signs are nonspecific, since they are seen in diabetes and other conditions.
Since 1990, there have been seven population-based epidemiologic studies (involving a total of 26,477 participants) of various signs of hypertensive retinopathy. In all seven studies, retinal photographs were used to define specific signs of retinopathy without regard to a predetermined grading system. All of the studies were conducted in the general community and included persons with and those without a history of hypertension.
In general, these studies show that signs of hypertensive retinopathy can be reliably identified with a standardized examination ofphotographs of the fundus. Reproducibility was substantial for the grading of retinal hemorrhages and microaneurysms (e.g., =0.80 to 0.99) and fair to moderate for the grading of arteriovenous nicking and focal arteriolar narrowing (=0.40 to 0.79). In four populations, generalized arteriolar narrowing was estimated from an assessment of retinal vessel diameters with the use of digitized photographs. This technique appears to have substantial reproducibility (i.e., the intraclass correlation coefficient ranged from 0.80 to0.99 in four studies).
On the basis of photographic grading, these epidemiologic studies show that signs of hypertensive retinopathy are common in people40 years of age or older, even in those without a history of hypertension. Prevalence rates ranged from 2 to 15 percent for various signs of retinopathy, in contrast to the earlier report from the Framingham Eye Study that found a prevalence of less than 1 percent among participants who underwent an ophthalmoscopic examination with dilation. The higher rates of prevalence in these more recent studies are probably due to a higher sensitivity of photography, as compared with clinical ophthalmoscopy, for detecting certain signs of retinopathy. However, there have been no studies that have directly compared the sensitivity or reliability of photography with that of ophthalmoscopy for the detection of hypertensive retinopathy, as there have been for diabetic retinopathy.
A higher prevalence of retinopathy has been reported among black persons than among whites, a difference that is explained in large part by the higher levels of blood pressure among blacks. The racial variation confirms the results of a previous population-basedsurvey that used direct ophthalmoscopy34 and suggests that retinal examination may be particularly useful for risk stratificationamong blacks. Variations in the prevalence of specific signs of hypertensive retinopathy according to age and sex have not been consistently demonstrated. There have been fewer studies of the incidence of hypertensive retinopathy. Two studies indicate that the incidence of various signs of retinopathy over a period of five to seven years ranges from 6 to 10 percent.
Numerous studies have confirmed the strong association between the presence of signs of hypertensive retinopathy and elevatedblood pressure. Two studies have further evaluated the effect of a history of elevated blood pressure on the occurrence of specific retinal signs. In both studies, generalized retinal arteriolar narrowing and arteriovenous nicking were associated with an elevation in blood pressure that had been documented six to eight years before the retinal assessment; the studies were controlled for concurrent blood-pressure levels. This association suggests that generalized narrowing and arteriovenous nicking are markers of vascular damage from chronic hypertension. In contrast, other signs (focal arteriolar narrowing, retinal hemorrhages, microaneurysms, and cotton-wool spots) were related to current but not previous blood-pressure levels and may therefore be more indicative of the severity of recent hypertension.
Furthermore, the observation of signs of retinopathy in people without a known history of hypertension suggests that these signs may be markers of a prehypertensive state. For example, generalized and focal narrowing of the retinal arterioles has been shown to predict the risk of hypertension in normotensive persons.38 Other factors unrelated to hypertension (e.g., hyperglycemia,inflammation, and endothelial dysfunction24) may also be involved in the pathogenesis of retinopathy.
The strongest evidence of the usefulness of an evaluation of hypertensive retinopathy for risk stratification is based on its association with stroke. It is well known that the retinal circulation shares anatomical, physiological, and embryologic features with the cerebral circulation. An autopsy study of patients with stroke showed a close correlation between retinal and cerebral arteriolar findings. Functional alterations in retinal blood flow in patients with lacunar stroke have also been reported.40
Epidemiologic data from four large, population-based studies showed independent associations between signs of hypertensiveretinopathy, as defined by the findings on retinal photographs, and the risk of stroke. The Atherosclerosis Risk in Communitiesstudy, a multisite cohort study, showed that some signs of retinopathy (e.g., retinal hemorrhages, microaneurysms, and cotton-woolspots) were associated with a risk of newly diagnosed clinical stroke that was two to four times as high as that for patients who did not have these signs, even when the analysis was controlled for the effects of long-term elevations in blood pressure, cigarettesmoking, elevated lipid levels, and other risk factors for stroke. This study has also shown that signs of retinopathy are associatedwith reduced cognitive performance on standardized neuropsychological tests, cerebral white-matter lesions, and cerebral atrophy as defined on the basis of findings on magnetic resonance imaging (MRI).
In the Atherosclerosis Risk in Communities study, the five-year relative risk of stroke among participants who had both hypertensiveretinopathy and cerebral lesions on MRI, as compared with those who had neither of these findings, was 18.1 (95 percent confidenceinterval, 5.9 to 55.4); among participants who had white-matter lesions only, the relative risk of stroke was 3.4 (95 percent confidence interval, 1.5 to 7.7). This pattern appears to reflect more severe or extensive subclinical cerebral microvascular disease in persons with both cerebral and retinal markers of hypertensive end-organ damage. In the Cardiovascular Health Study, after the analysis was controlled for elevated blood pressure and risk factors, persons with similar signs of retinopathy (retinal hemorrhages, microaneurysms, and cotton-wool spots) were twice as likely to have a history of stroke as were those who did not have these signs (odds ratio, 2.0; 95 percent confidence interval, 1.1 to 3.6). Population-based studies in Wisconsin and in Japanhave shown that the risks of fatal and nonfatal stroke are two to three times as high in persons with signs of retinopathy as they are in persons who do not have these signs — an association that is independent of cardiovascular risk factors.
These population-based studies also show substantially weaker and less consistent associations between other retinal changes (e.g., generalized and focal narrowing of the arterioles and arteriovenous nicking) and stroke, death from stroke, cognitive impairment, and cerebral changes on MRI. The retinal signs most strongly associated with stroke (i.e., retinal hemorrhages, microaneurysms, and cotton-wool spots) are correlated with a breakdown of the blood–retina barrier. The association of these signs with stroke may therefore suggest that disruption of the blood–brain barrier is a possible pathophysiological feature in the development of cerebrovascular disease. These findings also support the concept that an assessment of specific signs, rather than the presence or absence of hypertensive retinopathy, may be important for risk stratification.
There are fewer data regarding the association of hypertensive retinopathy and the risk of coronary heart disease. In the NationalHealth Examination Survey, persons with retinal arteriolar narrowing, as detected on ophthalmoscopy, were two to six times as likelyto have preexisting coronary heart disease as those without these changes, after the analysis was controlled for the presence or absence of hypertension and diabetes and for serum cholesterol levels. In a study of 560 men with hypertension and hyperlipidemia,the presence of hypertensive retinopathy predicted a doubling of the risk of coronary heart disease (relative risk, 2.1; 95 percent confidence interval, 1.0 to 4.2), and the presence of either generalized or focal narrowing of the arterioles predicted almost a tripling of this risk (relative risk, 2.9; 95 percent confidence interval, 1.3 to 6.2). In contrast, the Atherosclerosis Risk in Communities study showed that generalized narrowing of the retinal arterioles was associated with subsequent coronary heart disease in women (relative risk, 2.2; 95 percent confidence interval, 1.0 to 4.6) but not in men (relative risk, 1.1; 95 percent confidence interval, 0.7 to 1.8). This finding may reflect the higher risk of coronary microvascular disease among women than among men.
Some experimental studies and clinical trials have shown that signs of hypertensive retinopathy regress with the control of blood pressure, although spontaneous resolution of these signs in the presence of high blood pressure has also been reported.54 It is unclear whether antihypertensive medications that are thought to have direct beneficial effects on the microvascular structure (e.g., angiotension-converting–enzyme inhibitors) would reduce the damage of retinopathy beyond the reduction effected by lowered blood pressure. In a small study of 28 patients with mild hypertension who were randomly assigned to receive treatment with enalapril or hydrochlorothiazide, opacification of the retinal arteriolar wall was significantly reduced after 26 weeks of treatment with enalapril; no other signs of retinopathy were reduced.53 In contrast, hydrochlorothiazide did not have any effect on the signs of retinopathy. However, to date, there are no data from prospective, controlled trials that demonstrate that the specific reduction of hypertensiveretinopathy also reduces the morbidity and mortality associated with cardiovascular disease. It is also unclear whether the targeting of persons with hypertensive retinopathy for established risk-reducing interventions offers additional advantages over the use of strategies without regard to retinal findings.
The recent data suggest that there are several lines of future research. First, a standardized classification system for hypertensiveretinopathy should be developed that is relevant to contemporary clinical situations and reflects the recent data. Evidence from recent studies supports the development of a photographic classification system that would be similar to the photographic grading ofdiabetic retinopathy. However, it is not yet clear that retinal photography should be a routine part of the management of hypertensionor that photography is superior to ophthalmoscopy for the detection of signs of hypertensive retinopathy.
Second, additional prospective studies are needed that demonstrate independent associations of hypertensive retinopathy with various cardiovascular outcomes. For example, there are no recent studies focusing on whether signs of retinopathy predict otherhypertensive complications, such as renal dysfunction or congestive heart failure. It is also unclear whether a retinal examination would confer a greater benefit in specific subgroups of populations (e.g., younger people,45 women,49 and blacks33). An ongoinglongitudinal study involving a multiethnic population will provide further insights into these issues.54
Third, it is important to compare the relative value of a retinal assessment (based on an ophthalmoscopic examination performed with or without the use of photography) with other strategies of risk stratification (e.g., the use of electrocardiography and echocardiography). Finally, there is a need to evaluate whether specific therapy that is focused on the retinal microcirculation can reverse changes in retinopathy, and, if so, whether this approach will also ultimately result in a reduced cardiovascular risk.
How should the physician use the available evidence? This review provides compelling evidence that certain signs of hypertensiveretinopathy are associated with an increased cardiovascular risk that is independent of other risk factors. On the basis of the strength of these reported associations, we propose a simplified classification of hypertensive retinopathy — none, mild, moderate, and malignant — according to the severity of the retinal signs (Table 1). The physician may continue to provide routine care for patients with no retinopathy, undertake more vigilant monitoring of the cardiovascular risk in patients with mild retinopathy (i.e., those who have retinal arteriolar signs only), or adopt an aggressive approach to risk reduction in patients with moderate retinopathy (Figure 4). The few patients who have swelling of both optic disks and very high blood pressure (i.e., malignant retinopathy) need urgent antihypertensive treatment. In hypertensive patients with swelling of the optic disk, the physician should rule out anterior ischemic optic neuropathy, which occurs more frequently than malignant hypertensive retinopathy and is typically manifested as unilateral disk swelling, visual loss, and defects of the sectorial visual fields.
There is insufficient evidence to recommend a routine ophthalmoscopic consultation for all patients with hypertension. If the initialclinical findings are equivocal (e.g., there is borderline or inconsistent hypertension with no other evidence of target-organ damage), an ophthalmoscopic consultation may be useful to supplement the risk assessment and treatment decisions. For some patients (e.g., those with diabetes or visual symptoms), referral may also be important to rule out other conditions such as diabetic retinopathy or retinal-vein occlusion.
Signs of hypertensive retinopathy are common and are correlated with elevated blood pressure. Recent studies show that some of these signs (e.g., retinal hemorrhages, microaneurysms, and cotton-wool spots) predict stroke and death from stroke independentlyof elevated blood pressure and other risk factors. Patients with these signs of retinopathy may benefit from close monitoring of cerebrovascular risk and intensive measures to reduce that risk.
1. Burt VL, Cutler JA, Higgins M, et al. Trends in the prevalence, awareness, treatment and control of hypertension in the adultUS population: data from the health examination surveys, 1960 to 1991. Hypertension 1995;26:60-69. [Erratum, Hypertension1996;27:1192.] [Free Full Text]
2. The sixth report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure.Arch Intern Med 1997;157:2413-2446. [Erratum, Arch Intern Med 1998;158:573.] [Abstract]
3. Forette F, Seux ML, Staessen JA, et al. Prevention of dementia in randomised double-blind placebo-controlled SystolicHypertension in Europe (Syst-Eur) trial. Lancet 1998;352:1347-1351. [CrossRef][ISI][Medline]
5. Staessen JA, Gasowski J, Wang JG, et al. Risks of untreated and treated isolated systolic hypertension in the elderly: meta-analysis of outcome trials. Lancet 2000;355:865-872. [Erratum, Lancet 2001;357:724.] [CrossRef][ISI][Medline]
11. Collins R, Peto R, MacMahon S, et al. Blood pressure, stroke, and coronary disease. 2. Short-term reductions in blood pressure: overview of randomized drug trials in their epidemiological context. Lancet 1990;335:827-838. [CrossRef][ISI][Medline]
12. Psaty BM, Smith NL, Siscovick DS, et al. Health outcomes associated with antihypertensive therapies used as first-line agents: a systematic review and meta-analysis. JAMA 1997;277:739-745. [Abstract]
14. Sacks FM, Svetkey LP, Vollmer WM, et al. Effects on blood pressure of reduced dietary sodium and the Dietary Approaches to Stop Hypertension (DASH) diet. N Engl J Med 2001;344:3-10. [Free Full Text]
19. Appel LJ, Espeland MA, Easter L, Wilson AC, Folmar S, Lacy CR. Effects of reduced sodium intake on hypertensioncontrol in older individuals: results from the Trial of Nonpharmacologic Interventions in the Elderly (TONE). Arch Intern Med 2001;161:685-693. [Free Full Text]
21. Grimm RH Jr, Grandits GA, Cutler JA, et al. Relationships of quality-of-life measures to long-term lifestyle and drugtreatment in the Treatment of Mild Hypertension Study. Arch Intern Med 1997;157:638-648. [Abstract]
22. Hansson L, Zanchetti A, Carruthers SG, et al. Effects of intensive blood-pressure lowering and low-dose aspirin in patients with hypertension: principal results of the Hypertension Optimal Treatment (HOT) randomized trial. Lancet 1998;351:1755-1762. [CrossRef][ISI][Medline]
23. UK Prospective Diabetes Study Group. Tight blood pressure control and risk of macrovascular and microvascular complications in type 2 diabetes: UKPDS 38. BMJ 1998;317:703-713. [Erratum, BMJ 1999;318:29.] [Free Full Text]
24. SHEP Cooperative Research Group. Prevention of stroke by antihypertensive drug treatment in older persons with isolated systolic hypertension: final results of the Systolic Hypertension in the Elderly Program (SHEP). JAMA 1991;265:3255-3264. [Abstract]
25. Materson BJ, Reda DJ, Cushman WC, et al. Single-drug therapy for hypertension in men: a comparison of six antihypertensive agents with placebo. N Engl J Med 1993;328:914-921. [Erratum, N Engl J Med 1994;330:1689.] [Free Full Text]
26. Blood Pressure Lowering Treatment Trialists' Collaboration. Effects of ACE inhibitors, calcium antagonists, and other blood-pressure-lowering drugs: results of prospectively designed overviews of randomised trials. Lancet 2000;356:1955-1964. [CrossRef][ISI][Medline]
27. The Heart Outcomes Prevention Evaluation Study Investigators. Effects of an angiotensin-converting-enzyme inhibitor, ramipril, on cardiovascular events in high-risk patients. N Engl J Med 2000;342:145-153. [Erratum, N Engl J Med 2000;342:748, 1376.] [Free Full Text]
29. Estacio RO, Jeffers BW, Hiatt WR, Biggerstaff SL, Gifford N, Schrier RW. The effect of nisoldipine as compared with enalapril on cardiovascular outcomes in patients with non-insulin-dependent diabetes and hypertension. N Engl J Med 1998;338:645-652. [Free Full Text]
30. Hansson L, Lindholm LH, Ekbom T, et al. Randomised trial of old and new antihypertensive drugs in elderly patients: cardiovascular mortality and morbidity the Swedish Trial in Old Patients with Hypertension-2 study. Lancet 1999;354:1751-1756. [CrossRef][ISI][Medline]
31. Dahlof B, Devereux RB, Kjeldsen SE, et al. Cardiovascular morbidity and mortality in the Losartan Intervention For Endpoint reduction in hypertension study (LIFE): a randomised trial against atenolol. Lancet 2002;359:995-1003. [CrossRef][ISI][Medline]
32. Pahor M, Psaty BM, Alderman MH, et al. Health outcomes associated with calcium antagonists compared with other first-line antihypertensive therapies: a meta-analysis of randomised controlled trials. Lancet 2000;356:1949-1954. [CrossRef][ISI][Medline]
33. Major outcomes in high-risk hypertensive patients randomized to angiotensin-converting enzyme inhibitor or calcium channel blocker vs diuretic: the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT).JAMA 2002;288:2981-2997. [Free Full Text]
34. Wing LMH, Reid CM, Ryan P, et al. A comparison of outcomes with angiotensin-converting-enzyme inhibitors and diuretics for the treatment of hypertension in the elderly. N Engl J Med 2003;348:583-592. [Free Full Text]
35. Lewis EJ, Hunsicker LG, Bain RP, Rhode RD. The effect of angiotensin-converting-enzyme inhibition on diabetic nephropathy. N Engl J Med 1993;329:1456-1462. [Erratum, N Engl J Med 1993;330:152.] [Free Full Text]
37. Furberg CD, Psaty BM. Should calcium antagonists be first-line agents in the treatment of cardiovascular disease? The public health perspective. Cardiovasc Drugs Ther 1996;10:463-466. [ISI][Medline]
38. Saunders E, Weir MR, Kong BW, et al. A comparison of the efficacy and safety of a B-blocker, a calcium-channel blocker, and a converting enzyme inhibitor in hypertensive blacks. Arch Intern Med 1990;150:1707-1713. [Abstract]
41. Guidelines Subcommittee of the World Health Organization-International Society of Hypertension (WHO-ISH). 1999 World Health Organization-International Society of Hypertension Guidelines for the Management of Hypertension. J Hypertens 1999;17:151-183. [CrossRef][ISI][Medline]
46. Mann SJ, Blumenfeld JD, Laragh JH. Issues, goals, and guidelines for choosing first-line and combination antihypertensive drug therapy. In: Laragh JH, Brenner BM, eds. Hypertension: pathophysiology, diagnosis, and management, 2nd ed. Vol. 2. New York: Raven Press, 1995:2531-42.