INTRODUCTION — The majority of ischemic strokes are due to cardioembolism, large vessel atherothromboembolism, small vessel occlusive disease, or other unusual mechanisms. However, many ischemic strokes occur without a well-defined etiology and are labeled as cryptogenic.
This topic will provide an overview of cryptogenic stroke. A discussion of stroke classification and the clinical diagnosis of stroke subtypes is found separately. (See "Etiology, classification, and epidemiology of stroke" and "Clinical diagnosis of stroke subtypes".)
CLASSIFICATION — In its most useful clinical sense, the term cryptogenic stroke designates the category of ischemic stroke for which no probable cause is found despite a thorough diagnostic evaluation .
The cryptogenic stroke category was devised first, for research purposes, in the National Institute of Neurological Disorders and Stroke (NINDS) Stroke Data Bank [2,3] and later modified in the TOAST trial . Classification along these lines has become increasingly used in clinical practice, as optimal management relates to the underlying mechanism. (See "Etiology, classification, and epidemiology of stroke", section on 'TOAST classification'.)
By the TOAST classification (table 1), which is the one most commonly used in clinical practice, cryptogenic stroke (or stroke of undetermined origin in TOAST terminology) is defined as brain infarction that is not attributable to a source of definite cardioembolism, large artery atherosclerosis, or small artery disease despite a standard vascular, cardiac, and serologic evaluation. The category of stroke of undetermined etiology in the TOAST classification includes patients with less well-established potential causes of cardiac embolism, such as patent foramen ovale, aortic arch atheroma, and mitral valve strands, as well as potential prothrombotic disorders; stroke of undetermined etiology also includes patients with two or more equally plausible identified causes of stroke .
In addition to TOAST, there are several other ischemic stroke classification systems that include a category for stroke of undetermined cause, as discussed in detail separately (see "Etiology, classification, and epidemiology of stroke", section on 'SSS-TOAST and CCS classification'). Among these, the Causative Classification System (CCS) was designed to determine the most likely cause of stroke even when multiple possible mechanisms are present (table 2) [5,6].
Embolic stroke of undetermined source — Most cryptogenic strokes are likely embolic, as described below. This understanding is captured by a related concept, termed embolic stroke of undetermined source (ESUS), which is defined as a nonlacunar brain infarct without proximal arterial stenosis or cardioembolic sources . The concept of ESUS, moreover, implies that a full standard evaluation was done, whereas traditional definitions of cryptogenic stroke did not require a full evaluation. The criteria for ESUS are:
●Stroke detected by CT or MRI that is not lacunar (lacunar is defined as a subcortical infarct in the distribution of the small, penetrating cerebral arteries whose largest dimension is ≤1.5 cm on CT or ≤2.0 cm on MRI diffusion images)
●Absence of extracranial or intracranial atherosclerosis causing ≥50 percent luminal stenosis of the artery supplying the area of ischemia
●No major-risk cardioembolic source of embolism (ie, no permanent or paroxysmal atrial fibrillation, sustained atrial flutter, intracardiac thrombus, prosthetic cardiac valve, atrial myxoma or other cardiac tumors, mitral stenosis, recent (within four weeks) myocardial infarction, left ventricular ejection fraction <30 percent, valvular vegetations, or infective endocarditis)
●No other specific cause of stroke identified (eg, arteritis, dissection, migraine, vasospasm, drug abuse)
Thus, ESUS represents a subset of cryptogenic stroke and emphasizes the likelihood that most strokes of unexplained etiology are probably embolic from an unestablished source [1,7].
POSSIBLE MECHANISMS — Numerous mechanisms for cryptogenic stroke have been proposed. The most likely mechanisms include the following:
●Cardiac embolism secondary to occult paroxysmal atrial fibrillation (AF), aortic atheromatous disease, or other cardiac sources
●Paradoxical embolism, which originates in the systemic venous circulation and enters the systemic arterial circulation through a patent foramen ovale (PFO), atrial septal defect, ventricular septal defect, or extracardiac communication such as a pulmonary arteriovenous malformation
●Undefined thrombophilia (ie, hypercoagulable states including those related to antiphospholipid antibodies or to occult cancer with hypercoagulability of malignancy)
●Substenotic cerebrovascular disease (ie, intracranial and extracranial atherosclerotic disease causing less than 50 percent stenosis) and other vasculopathies (eg, dissection)
It is also likely that important but unidentified mechanisms exist, awaiting discovery.
Embolism from occult sources in the heart or aorta — The embolic appearance of most cryptogenic strokes implies that the cause is embolism from an occult source in the heart, aorta, or large artery. Cardioaortic conditions with a low or uncertain risk for embolic stroke include difficult to diagnose ("occult") or subclinical atrial fibrillation and related atrial cardiopathies, atrial septal abnormalities, complex aortic atheroma, and others (table 3).
Occult atrial fibrillation — Occult paroxysmal AF refers to asymptomatic paroxysmal AF in a patient without a prior history of AF, which is detected only by monitoring techniques. Evidence linking occult AF and cryptogenic stroke comes from the prospective ASSERT study of 2580 subjects, age ≥65 years, with hypertension and no history of AF who had recent implantation of a pacemaker or defibrillator . At three months, subclinical atrial tachyarrhythmias detected by the implanted devices had occurred in 10 percent of patients, and were associated with an increased risk (at a mean of 2.5 years) for clinical AF (hazard ratio [HR] 5.6, 95% CI 3.8-8.2) and for the combined endpoint of ischemic stroke or systemic embolism (HR 2.5, 95% CI 1.3-4.9). Among subjects with at least three months of continuous monitoring who experienced ischemic stroke or systemic embolism (n = 51), subclinical AF was detected overall in 26 (51 percent) . However, subclinical AF occurring ≤30 days before ischemic stroke or systemic embolism was detected in only 4 subjects (8 percent). Thus, while subclinical AF was associated with an increased risk of embolic events, there was no definite temporal relationship of subclinical AF with stroke in most subjects.
Atrial septal abnormalities — Atrial septal abnormalities, including PFO, atrial septal aneurysm, and atrial septal defect have been associated with cryptogenic stroke, although the strength and clinical significance of this association is uncertain. There is an increased prevalence of PFO and atrial septal aneurysm in patients who have had a cryptogenic stroke, particularly in patients <55 years old. However, population-based studies suggest that PFO and large PFO are not independent risk factors for stroke. In addition, prospective data suggest that PFO alone is not associated with a meaningfully increased risk of recurrent stroke or death in patients who have already had a cryptogenic stroke. These issues are discussed in detail separately. (See "Atrial septal abnormalities (PFO, ASD, and ASA) and risk of cerebral emboli in adults", section on 'Risk of embolic stroke'.)
Atrial cardiopathies — Even in the absence of diagnosed atrial fibrillation, biomarkers of atrial dysfunction are associated with risk of ischemic stroke. For example, serum troponin and pro-brain natriuretic peptide (proBNP) are associated with both AF and stroke. (See "Blood biomarkers for stroke", section on 'BNP and cardioembolic stroke'.)
Similarly, P wave terminal force velocity in lead V1 (PWTFV1), a measure of atrial contraction that can be measured on the electrocardiogram, is associated with stroke risk even in the absence of AF . Although cardioembolism is presumed to be the most likely mechanism of stroke in patients with elevated proBNP or PWTFV1, it is difficult to establish a cause and effect relationship between these elevated cardiac biomarkers and occult cardioembolism; the association is confounded because cardiac disease and elevated cardiac biomarkers are also markers of systemic atherosclerosis. In addition, these biomarkers are not widely available in clinical practice and their utility for management is still uncertain. However, these biomarkers have the advantage of being measurable at the time of stroke without the need for long-term monitoring, and thus provide the potential to detect a high risk of cardioembolism. (See "Blood biomarkers for stroke", section on 'BNP and cardioembolic stroke'.)
Further prospective study is needed to confirm that any of these biomarkers reliably predict a cardioembolic stroke mechanism and response to anticoagulant therapy in secondary stroke prevention.
Aortic embolism — Thoracic aortic atherosclerotic plaques are an important potential source of systemic emboli, leading to stroke, transient ischemic attack, and embolization to other arterial beds. The risk of thromboembolism in patients with aortic atherosclerosis is increased when there is complex plaque, which is defined as thickness >4 mm or ulceration. (See "Embolism from aortic plaque: Thromboembolism", section on 'Complex aortic plaque'.)
Besides proximal aortic atheromas, distal aortic sources of embolism have been proposed as a potential cause for cryptogenic stroke. One study using cardiac MRI suggested that complex atheromas in the descending aortic arch could lead to stroke via retrograde flow . During diastole, retrograde flow in the descending aorta reached the great vessels supplying the brain in up to 24 percent of patients with cryptogenic stroke. This finding suggests that embolic material in the descending aortic arch could enter the cerebral vasculature during retrograde flow and cause ischemic stroke.
Pulmonary shunts — Intrapulmonary right-to-left shunts due to pulmonary arteriovenous malformations or arteriovenous fistulas have been associated with cryptogenic stroke in several small studies [12-16]. (See "Pulmonary arteriovenous malformations: Clinical features and diagnostic evaluation in adults", section on 'Neurologic'.)
Other potential causes of stroke include coarctation of the aorta and aortic dissection.
The relationship of aortic embolism and stroke is reviewed in detail separately. (See "Etiology, classification, and epidemiology of stroke", section on 'Aortic atherosclerosis' and "Embolism from aortic plaque: Thromboembolism".)
Substenotic atherosclerotic disease — Some cases of cryptogenic stroke may be caused by undetected large vessel disease, including occult atherosclerosis and nonstenosing, unstable plaques.
●In 37 patients with stroke of no determined cause who were followed prospectively for one year in a South Korean study, 11 patients suffered recurrent ischemic stroke . After stroke recurrence, six patients had evidence of large artery atherosclerosis, one was classified as cardioembolic (new onset atrial fibrillation), and four were still cryptogenic. The authors proposed that in their population, the predominant mechanism of cryptogenic stroke was progression of large vessel atherosclerosis or plaque instability.
●Data from autopsy studies suggest that ischemic stroke can be associated with lesser degrees of extracranial and intracranial large vessel stenosis (eg, 30 to 70 percent) or with vulnerable plaques without appreciable luminal compromise. Supporting evidence comes from the following observations:
•In a case-control study that included 259 patients with fatal ischemic stroke, intracranial atherosclerotic plaques (with or without stenosis) were noted in 62 percent . Furthermore, plaques with superimposed thrombi and stenosis of 30 to 70 percent were considered responsible for infarcts in four cases (1.5 percent), a group that would have been classified as cryptogenic in non-autopsy studies .
•In a subsequent study from the same investigators, plaques and stenoses involving the origin or proximal vertebral artery were present in more than twice as many patients with infarcts in posterior circulation as compared with anterior circulation infarcts (adjusted OR 2.10, 95% CI 1.01-4.38) . These lesions may be responsible for a larger proportion of strokes in the brainstem and posterior circulation than previously appreciated.
Other causes — It is possible that subtle or undetected abnormalities of the larger arteries, coagulation system, and genetic factors may be missed during the initial evaluation. These conditions include:
●Nonatherosclerotic vasculopathies, such as dissection, fibromuscular dysplasia, reversible cerebral vasoconstriction syndromes (RCVS), and vasculitis
●Occult hypercoagulable states, such as the antiphospholipid syndrome, genetic thrombophilia, and hypercoagulable state associated with malignancy
●Rare genetic conditions may present with stroke in the young; monogenic syndromes associated with an increased risk of ischemic stroke include Fabry disease, cerebral autosomal recessive arteriopathy with subcortical infarctions and leukoencephalopathy (CADASIL), and sickle cell disease
However, detection of any of the above conditions would alter the diagnostic classification of stroke from cryptogenic stroke to stroke of other determined cause.
EPIDEMIOLOGY AND RISK FACTORS — Large epidemiologic studies have consistently reported that cryptogenic stroke accounts for 25 to 40 percent of ischemic stroke [20-28]. The incidence and prevalence of stroke subtypes among these studies may vary based upon the demographics of the study population, diagnostic definitions, extent of diagnostic evaluation, and methodology. Thus, it is conceivable that some strokes of other determined cause (eg, migraine, dissection, vasculitis) were misclassified in the undetermined category (ie, as cryptogenic) due to inadequate work-up or the limitations of diagnostic detection. However, given the rarity of these other causes in most registries (usually representing less than 5 percent of all strokes), this would not account for all cryptogenic strokes.
Demographic factors — The risk of cryptogenic stroke may vary by demographics, with higher incidence rates in blacks and Hispanics than in whites, but no clear association has been found for age and gender.
With the exception of the strokes classified in TOAST as "other determined etiology" (which includes dissection), all stroke subtypes are rare in the young, and incidence rates rise dramatically with increasing age. A few studies have reported that cryptogenic stroke disproportionately affects younger individuals, but the evidence is inconsistent.
●In the Northern Manhattan Stroke Study (NOMASS, 1993 to 1996), 55 percent of strokes in the young (age <45 years) were cryptogenic versus 42 percent in the older (age >45 years) group .
●In a 2003 meta-analysis, young age (defined as <50 years) was inversely associated with cryptogenic stroke with a total OR of 0.6 (95% CI 0.4-1.0, p = 0.05) .
●Other stroke registries found lower rates (23 to 34 percent) in younger age groups, which were similar to those in older age groups [30-32].
The incidence of cryptogenic stroke may be higher in blacks and Hispanics than in whites. In NOMASS, incidence rates of all ischemic stroke subtypes, including cryptogenic stroke, were higher in blacks and Hispanics than in whites . In the Greater Cincinnati/Northern Kentucky Stroke Study (GCNKSS), blacks had twice the annual incidence rate of cryptogenic stroke as whites (125 versus 65 per 100,000 persons), a result not confounded by differential testing patterns among black versus white patients . In San Diego, an increased prevalence (nearly 46 percent) of cryptogenic stroke was seen in Mexican-American patients, a statistic again that was not explained by differences in diagnostic testing .
Other risk factors — Although risk factors often help unravel stroke mechanisms, the two are not always synonymous. While it is possible to compare risk factors for cryptogenic stroke versus other stroke subtypes, the comparison is hindered in large part by definitional constraints. As an example, atrial fibrillation will be rare in cryptogenic stroke because of the way in which the subtypes are defined. In addition, risk factors that are associated with large artery ischemic stroke (eg, hypertension, hyperlipidemia, peripheral vascular disease, and diabetes mellitus) and cardioembolic stroke (eg, acute coronary events) are underrepresented in patients with cryptogenic stroke .
Several studies have documented that hypertension is less common in cryptogenic stroke compared with other stroke subtypes [23,24,26,27,34,36]. However, patients with cryptogenic stroke may have an increased prevalence of hypertension compared with stroke-free controls, and one case-control study found that hypertension was associated with cryptogenic stroke (OR 4.5, 95% CI 1.5-13.2) .
The prevalence of cardiac disease among patients with cryptogenic stroke varies from 10 to 30 percent. In Rochester, coronary artery disease was less common in the undetermined (ie, cryptogenic) subtype than in the large artery atherosclerosis subtype .
Studies assessing the prevalence of prothrombotic states and genetic polymorphisms predisposing individuals to thrombosis have not yielded convincing evidence that these are more common in patients with cryptogenic stroke than nonstroke controls [38-41]. Nevertheless, the available reports are small and not definitive.
CLINICAL FEATURES — Patients with cryptogenic stroke typically present with sudden onset of focal neurologic deficits and an embolic infarct topography on brain imaging. Superficial hemispheric infarction is present in 62 to 84 percent of patients [20,42]. In one study of patients with cryptogenic stroke, cortical signs were present in 27 percent, and abrupt onset occurred in 59 percent . Lacunar syndromes are rare, accounting for usually less than 5 percent [20,42]. The severity of the initial presentation varies but, on average, tends to be milder than cardioembolic strokes and worse than lacunar strokes [17,27,42-44].
A significant proportion of cryptogenic strokes adhere to embolic infarct topography on brain imaging (CT or MRI). Forty percent of cryptogenic strokes in the Stroke Data Bank were found to have cortical infarcts . Among 314 patients with cryptogenic stroke in the PFO-ASA study, 56 percent had superficial infarcts . The German Stroke Study found that parenchymal hemorrhagic transformation occurred in approximately 2 percent of patients with stroke of unknown etiology in the first seven days, comparable to the percentage among cardioembolic stroke, suggesting an embolic mechanism . Large subcortical strokes (>15 mm) also tend to be either cryptogenic or cardioembolic in origin.
Noninvasive vascular imaging typically does not identify a relevant vessel stenosis or occlusion in most patients with cryptogenic stroke, except the possibility of distal branch occlusion. However, in prospective data from the Stroke Data Bank, conventional angiography revealed large artery lesions in a significant minority of patients initially classified as having infarcts of undetermined cause .
In patients with cryptogenic stroke, the most frequently found abnormalities with echocardiography are patent foramen ovale (PFO), atrial septal aneurysm (ASA), and aortic atheromas [45,46]. The timing of the transesophageal echocardiography (<72 hours versus >72 hours) in relation to the index stroke does not appear to alter sensitivity . The clinical significance of most of these findings is still unclear, with conflicting studies on the relative risks and appropriate management.
EVALUATION AND DIAGNOSIS — Cryptogenic stroke is a diagnosis of exclusion based upon a thorough investigation for potential stroke etiologies. The diagnosis of cryptogenic stroke is made when a standard evaluation (see 'Standard evaluation' below) reveals no probable cause; there is no definite evidence of cardioembolism, large artery atherosclerosis, small artery disease, or other determined etiology, and no evidence of atrial fibrillation on a 12-lead electrocardiogram (ECG) or on 24-hour cardiac monitoring.
Patient age influences the relative likelihood of possible ischemic stroke mechanisms . Cervicocephalic artery dissection is the most common cause in young adults; other considerations include congenital cardiac defects, recent pregnancy, hypercoagulable states, illicit drug use, metabolic disorders, and migraine. (See "Ischemic stroke in children and young adults: Etiology and clinical features", section on 'Etiologies and risk factors in young adults'.)
Premature atherosclerosis and acquired cardiac disease are increasingly prevalent in adults older than 30 years of age, and occult atrial fibrillation is increasingly common in patients older than 60 years of age .
Standard evaluation — The standard evaluation of patients with acute ischemic stroke includes a history and physical examination, brain imaging to determine the location and topography of the lesion, and vessel imaging and a cardiac evaluation to help determine the most likely cause. Laboratory testing typically includes a complete blood count, platelet count, cardiac enzymes and troponin, prothrombin time, international normalized ratio (INR), and activated partial thromboplastin time.
Additional studies can be pursued if the standard evaluation fails to determine the probable cause. (See "Initial assessment and management of acute stroke" and "Overview of the evaluation of stroke".)
Brain imaging — Urgent brain imaging with CT or MRI is mandatory in all patients with sudden neurologic deterioration or acute stroke (see "Neuroimaging of acute ischemic stroke"). Brain MRI with diffusion-weighted imaging is superior to noncontrast CT for the detection of acute ischemia, small infarcts, and infarcts located in the brainstem. The localization, topography, and distribution of ischemic brain lesions on MRI and CT can suggest a specific stroke mechanism [1,48]:
●Isolated superficial cerebral or cerebellar infarction suggests an embolic mechanism from a large artery, heart, or aorta
●Cortical or large subcortical infarcts in multiple vascular territories suggest a proximal source of embolism from the heart or aorta
●Infarcts of varying age in a single vascular territory suggest a large artery source of embolism
●Infarcts along the boundary regions between the major cerebral arteries (ie, border zone or watershed regions) suggest the stroke mechanism is low flow (hypoperfusion) or multiple small emboli
●Small subcortical infarcts suggest lacunar infarction from small vessel disease
The diagnosis of small vessel disease as the cause of ischemic stroke is generally confirmed by neuroimaging when the location of a small noncortical infarct on CT or MRI correlates with the clinical features of a lacunar stroke syndrome. However, a small deep infarct may be considered cryptogenic when found in a patient <50 years of age with no standard vascular risk factors and no white matter hyperintensities or prior small deep infarcts [1,49].
Vessel imaging — Vessel imaging to identify the lesion (eg, atherosclerotic stenosis or occlusion, dissection) responsible for stroke can be done with magnetic resonance angiography (MRA), computed tomography angiography (CTA), carotid duplex ultrasonography and transcranial Doppler ultrasonography, or conventional angiography (see "Neuroimaging of acute ischemic stroke"). Neurovascular imaging should assess the extracranial (internal carotid and vertebral) and intracranial (internal carotid, vertebral, basilar, and Circle of Willis) large vessels.
Noninvasive methods are generally used unless urgent endovascular therapy is planned. MRA or CTA is preferred, while the combination of ultrasound methods (duplex and transcranial Doppler) can be used if CTA and MRA are unavailable or contraindicated. Availability and expertise at individual centers is a major factor in the choice of the initial noninvasive neurovascular studies.
Various neuroimaging modalities may be used to confirm a diagnosis of dissection, but fat-saturated T1 MRI is capable of revealing the intramural hematoma caused by dissection in vessels that otherwise have a normal appearance on MRA and CTA. (See "Spontaneous cerebral and cervical artery dissection: Clinical features and diagnosis", section on 'Neuroimaging'.)
Conventional angiography is usually reserved for situations where acute intraarterial intervention is being considered and for follow-up when noninvasive studies are inconclusive.
Cardiac and aortic evaluation — The basic cardiac evaluation of acute ischemic stroke includes an electrocardiogram, cardiac monitoring for at least the first 24 hours after stroke onset to look for occult atrial fibrillation (AF), and echocardiography. (See "Overview of the evaluation of stroke", section on 'Cardiac evaluation'.)
Both transthoracic echocardiography (TTE) and transesophageal echocardiography (TEE) are effective diagnostic tests for the evaluation of suspected cardioaortic source of embolism. In most patients, TEE yields higher quality images and has a greater sensitivity and specificity than TTE, but a few conditions (eg, left ventricular thrombus) are better seen on TTE. However, TEE is an uncomfortable invasive procedure that may not be tolerated by very ill patients. Because it is less invasive and readily available in most institutions, TTE is often reasonable as the initial test of choice (see "Echocardiography in detection of cardiac and aortic sources of systemic embolism").
TTE is the preferred initial test for the majority of patients with a suspected cardiac or aortic source of emboli, including:
●Patients ≥45 years
●Patients with a high suspicion of left ventricular thrombus
●Patients in whom TEE is contraindicated (eg, esophageal stricture, unstable hemodynamic status) or who refuse TEE
TEE may be especially helpful to localize the source of embolism in the following circumstances:
●Patients <45 years without known cardiovascular disease (ie, absence of myocardial infarction or valvular disease history)
●Patients with a high pretest probability of a cardiac embolic source in whom a negative TTE would be likely to be falsely negative
●Patients with atrial fibrillation and suspected left atrial or left atrial appendage thrombus, especially in the absence of therapeutic anticoagulation, but only if the TEE would impact management
●Patients with a mechanical heart valve
●Patients with suspected aortic pathology
For patients ≤60 years of age with an embolic-appearing cryptogenic stroke or TIA, particularly those who lack cardiovascular risk factors, we suggest TEE when TTE is nondiagnostic. The TEE should be performed with color Doppler study and agitated saline contrast injection at rest, with cough, and Valsalva maneuver. Although data are limited, a prospective study of 61 patients with embolic stroke of undetermined source (ESUS) found that abnormalities on TEE changed the therapeutic strategy in 16 percent .
The diagnosis of stroke or TIA due to paradoxical embolism through a patent foramen ovale (PFO) or atrial septal defect is usually one of exclusion; a PFO or atrial septal defect is considered a potential cause of cryptogenic embolic stroke or TIA in patients who are ≤60 years of age with no other identifiable cause. In presence of a PFO or atrial septal defect, it is reasonable to search for a source of thrombus in the leg veins with Doppler of the lower extremities as standard test and obtain a hypercoagulable panel in those <45 years of age. Pelvic magnetic resonance venography is of limited utility but could be used in specific scenarios (eg, recent pelvic surgery or mass) [51,52].
Advanced evaluation — Additional testing is warranted for patients with ischemic stroke when the cause is undetermined despite a standard evaluation described above.
●Prolonged cardiac monitoring – Paroxysmal atrial fibrillation, if transient, infrequent, and largely asymptomatic, may be undetected on standard cardiac monitoring such as continuous telemetry and 24- or 48-hour Holter monitors. The optimal monitoring method (ie, continuous telemetry, ambulatory electrocardiography, serial ECG, transtelephonic ECG monitoring, or implantable loop recorders) is uncertain, though longer durations of monitoring are likely to obtain the highest diagnostic yield. We suggest ambulatory cardiac monitoring for several weeks (eg, 30 days) for all adult patients with a cryptogenic ischemic stroke or cryptogenic TIA. (See "Overview of the evaluation of stroke", section on 'Monitoring for occult atrial fibrillation'.)
●Advanced cardiac imaging – Cardiac structural imaging with MRI can be helpful for identifying potential sources of embolism that may be missed by echocardiography, including left ventricular thrombi, isolated left ventricular noncompaction, and complex aortic atheroma [48,53]. (See "Clinical utility of cardiovascular magnetic resonance imaging" and "Isolated left ventricular noncompaction" and "Embolism from aortic plaque: Thromboembolism".)
●Vascular studies – Advanced vascular imaging can be useful for demonstrating lesions that escape detection on standard MRA and CTA. These are considered in specific scenarios such as small vessel vasculitis or vasculopathy (eg, catheter angiography) or subclinical atherosclerotic plaques (eg, high-resolution MRA).
Conventional angiography is superior to standard noninvasive methods (MRA, CTA, and ultrasonography) for visualizing small and medium sized arteries. Digital subtraction angiography, the most widely used method of conventional catheter-based angiography, remains the gold standard for determining the degree of arterial stenosis and for identifying nonatherosclerotic vasculopathy and vasculitis (see "Neuroimaging of acute ischemic stroke", section on 'Conventional angiography'). The yield of catheter angiography may be highest in the first hours after stroke onset, since vascular abnormalities may resolve in the acute phase .
Monitoring with transcranial Doppler (TCD) ultrasonography for 30 to 60 minutes may be useful to detect asymptomatic microemboli arising from the heart, aorta, or large arteries, and thereby point to the possible embolic source of the cryptogenic stroke. (See "Management of asymptomatic carotid atherosclerotic disease", section on 'Asymptomatic embolism'.)
Advanced, high-resolution MRI techniques allow direct visualization of the vessel wall, rather than just luminal narrowing as detected by conventional imaging . These methods show promise for the evaluation of intracranial arterial pathology, such as differentiating atherosclerotic, vasospastic, and inflammatory vasculopathies, demonstrating nonstenotic plaques that occlude penetrating arteries, and identifying features that suggest plaque vulnerability [55-59]. However, high-resolution MRI has limited availability and requires further study to establish reliability and to determine how well imaging findings correlate with vessel pathology .
●Hematologic testing – Hematologic testing for arterial hypercoagulable states (eg, antiphospholipid syndrome and hyperhomocysteinemia) is indicated for many patients with cryptogenic stroke, particularly for patients who are young, have a history of lupus or symptoms compatible with lupus, or have features suggestive of antiphospholipid syndrome such as unexplained venous or arterial thrombotic events, miscarriages, or unexplained thrombocytopenia. (See "Clinical manifestations of antiphospholipid syndrome" and "Overview of homocysteine", section on 'Hyperhomocysteinemia'.)
In addition to testing for the antiphospholipid syndrome, additional testing for hypercoagulable states associated with venous thrombosis (eg, Factor V Leiden mutation, prothrombin gene mutation, protein S deficiency, protein C deficiency, and antithrombin deficiency) is suggested by some experts for patients with evidence of a cardiac or pulmonary right-to-left shunt .
For patients with cryptogenic stroke and systemic or constitutional symptoms suggestive of vasculitis, screening tests include erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), serum cryoglobulins, antinuclear antibody (ANA), antineutrophil cytoplasmic antibody (ANCA), and complement levels. (See "Overview of and approach to the vasculitides in adults".)
Specialized evaluation — In some patients with recurrent cryptogenic stroke in whom standard and advanced evaluations are nondiagnostic, a search for other rare genetic and acquired risk factors may be indicated .
Specialized testing may include the following investigations:
●Testing for occult malignancy with mammography, stool Hemoccult, and CT of the chest, abdomen, and pelvis.
●Testing for genetic syndromes associated with stroke, including genes responsible for Fabry disease, mitochondrial encephalomyopathy with lactic acidosis and stroke-like episodes (MELAS), cerebral autosomal recessive arteriopathy with subcortical infarctions and leukoencephalopathy (CADASIL), and others.
●A lumbar puncture with cerebral spinal fluid analysis for patients with symptoms suggestive of primary angiitis of the central nervous system (PACNS), such as unexplained TIA or stroke (often multiple strokes in different vascular territories), headache, spinal cord dysfunction, or cognitive impairment. (See "Primary angiitis of the central nervous system in adults".)
●A brain biopsy, which is sometimes diagnostic for patients with suspected vasculitis, intravascular lymphoma, or infectious causes.
●Studies to detect a pulmonary arteriovenous malformation, a rare cause of ischemic stroke, which may be suspected in patients who have features such as a nodule on chest radiography, stigmata of a right-to-left shunt (eg, cyanosis, clubbing, history of ischemic stroke or brain abscess), unexplained hemoptysis, hypoxemia or dyspnea, and patients with suspected or known hereditary hemorrhagic telangiectasia. A delayed right-to-left shunt is often detected on transthoracic echocardiography with contrast (ie, bubble study) and the diagnosis can be confirmed with chest CT or pulmonary angiography. (See "Pulmonary arteriovenous malformations: Clinical features and diagnostic evaluation in adults".)
TREATMENT — Acute therapy for patients with cryptogenic stroke is no different from other types of ischemic stroke (see 'Acute therapy' below), while the choice of antithrombotic therapy for secondary prevention is challenging because no clear treatment target can be identified. (See 'Secondary prevention' below.)
Acute therapy — There are no convincing data that acute ischemic stroke management should differ by stroke subtype . Intravenous thrombolysis with tPA (alteplase) is beneficial for eligible patients with ischemic stroke who can be treated within 4.5 hours of stroke onset, and mechanical thrombectomy using a second-generation stent retriever device is beneficial for patients with ischemic stroke caused by a large artery occlusion in the proximal anterior circulation. Acute management for patients with cryptogenic stroke who are not eligible for these interventions is also similar to patients with other ischemic stroke subtypes. (See "Initial assessment and management of acute stroke" and "Approach to reperfusion therapy for acute ischemic stroke" and "Mechanical thrombectomy for acute ischemic stroke" and "Antithrombotic treatment of acute ischemic stroke and transient ischemic attack".)
Secondary prevention — For secondary prevention, most patients with an ischemic stroke or transient ischemic attack (TIA) should be treated with all available risk reduction strategies. Currently viable strategies include blood pressure reduction, antithrombotic therapy, statin therapy, and lifestyle modification. (See "Overview of secondary prevention of ischemic stroke".)
The choice of antithrombotic therapy for secondary stroke prevention after cryptogenic stroke is challenging because no clear treatment target can be identified, with the exception of a patent foramen ovale (PFO) with right-to-left shunt (see 'Presence of a PFO' below). Nevertheless, antiplatelet therapy is recommended for most patients with noncardioembolic stroke, including cryptogenic stroke (algorithm 1) [61,62]. (See "Antiplatelet therapy for secondary prevention of stroke".)
There is a high degree of uncertainty regarding the optimal management of patients with cryptogenic stroke who have an isolated atrial septal aneurysm (ASA), or atheromatous aortic disease. The optimal management of specific coagulation disorders is also unclear at the moment. Therefore, antiplatelet therapy is usually recommended for patients with cryptogenic stroke who have these conditions (algorithm 1) [61,62].
Pending long-term cardiac monitoring — Although there is no consensus among experts, we recommend antiplatelet therapy while awaiting the results of long-term cardiac monitoring in patients with a first cryptogenic stroke, and continuing antiplatelet therapy if no atrial fibrillation is detected on long-term monitoring.
Some stroke experts use anticoagulation when there is a high suspicion for a cardiac source of embolism despite the lack of evidence from randomized trials to support such an approach. For example, while awaiting the results of long-term cardiac monitoring, some experts would start empiric oral anticoagulation at hospital discharge for patients with acute embolic stroke that is cryptogenic after standard evaluation if there are multiple risk factors for occult atrial fibrillation . These include higher CHA2DS2-VASc score (table 4), the presence of cortical or large subcortical infarcts in multiple vascular territories, and evidence of left atrial cardiopathy (eg, left atrial dilatation, strain, reduced emptying fraction, left atrial appendage size and single lobe morphology, P wave dispersion on electrocardiogram (ECG), and frequent atrial premature beats) . Further antithrombotic treatment is directed by the presence or absence of atrial fibrillation detected on 30-day cardiac monitoring, and the frequency and duration of atrial fibrillation if detected.
There is no proven benefit of anticoagulation compared with antiplatelet therapy for preventing recurrent ischemic stroke in patients with cryptogenic stroke.
●The NAVIGATE-ESUS trial randomly assigned patients with embolic stroke of undetermined source (ESUS) to treatment with rivaroxaban or aspirin . The trial was stopped early for futility after an interim analysis showed no benefit of rivaroxaban on the rate of stroke or systemic embolism but an increase in major bleeding in the rivaroxaban arm . Therefore, rivaroxaban should not be used as empiric treatment for patients with cryptogenic stroke, including ESUS.
●The Warfarin-Aspirin Recurrent Stroke Study (WARSS) compared aspirin with warfarin in the prevention of recurrent ischemic stroke among noncardioembolic stroke patients and found no superiority of warfarin over aspirin . Among patients with cryptogenic stroke, the event rate (recurrent stroke or death) at two years was not significantly different for the warfarin-treated group compared with the aspirin-treated group (15.0 versus 16.5 percent, respectively).
●A post-hoc analysis of WARSS data showed that warfarin therapy was associated with significantly fewer recurrent strokes or deaths at two years compared with aspirin in selected subgroups of patients with cryptogenic stroke: those with mild stroke severity (National Institutes of Health Stroke Scale score ≤5), those with posterior circulation infarcts sparing the brainstem, and those with no hypertension at baseline . In the subgroup of patients with cryptogenic stroke who had an infarct topography consistent with an embolic mechanism, the event rate was lower with warfarin compared with aspirin (12 versus 18 percent, hazard ratio 0.66, 95% CI 0.37-1.15), but this difference did not achieve statistical significance. In another post-hoc analysis of WARSS data, there was a significant reduction in the composite end point of stroke or death favoring warfarin over aspirin treatment among patients with highly elevated levels of N-terminal pro-brain natriuretic peptide (NT-proBNP), a marker associated with atrial fibrillation and cardiac dysfunction . Since these results come from post-hoc analyses based on relatively small numbers of patients, they must be interpreted with great caution, and further prospective studies are needed to determine if warfarin is beneficial in specific subgroups of patients with cryptogenic stroke.
●The ARCADIA trial will test whether apixaban is superior to standard dose aspirin in preventing recurrent stroke among patients with unexplained ischemic stroke and biomarkers indicative of atrial cardiopathy .
Occult or subclinical atrial fibrillation — We suggest anticoagulant therapy with warfarin or a non-vitamin K antagonist oral anticoagulant (NOAC) for patients initially diagnosed with cryptogenic stroke who have atrial fibrillation of any duration detected on long-term monitoring, even if detected remotely from the incident stroke.
Most experts agree that occult or subclinical atrial fibrillation found on long-term monitoring should be treated with anticoagulants. However, there is no consensus regarding the use of anticoagulant treatment for patients when monitoring detects only very brief (eg, ≤30 seconds) or rare episodes of paroxysmal atrial fibrillation.
Presence of a PFO — Percutaneous PFO closure in addition to antiplatelet therapy is suggested for patients age ≤60 years with an embolic-appearing cryptogenic ischemic stroke (ie, no evident source of stroke despite a comprehensive evaluation) who have a PFO with a right-to-left shunt detected by bubble study. An exception is that patients with cryptogenic stroke and PFO who have acute deep venous thrombosis, pulmonary embolism, or other venous thromboembolism are generally treated with anticoagulation for at least several months. The management of patients with cryptogenic stroke and PFO is discussed in greater detail separately. (See "Treatment of patent foramen ovale (PFO) for secondary stroke prevention".)
Recurrent cryptogenic stroke — For patients on antiplatelet therapy who have a recurrent cryptogenic stroke and no atrial fibrillation on re-evaluation with long-term cardiac monitoring, options include continuing the same antiplatelet agent or switching to another antiplatelet agent; for patients with recurrent embolic stroke of undetermined source (see 'Embolic stroke of undetermined source' above), switching to empiric anticoagulant therapy is also a reasonable option.
PROGNOSIS — Compared with other stroke subtypes, cryptogenic stroke tends to have a better prognosis at three months, six months, and one year. Approximately 50 to 60 percent of patients score <2 on the modified Rankin Scale (table 5) at follow-up [17,27,44,69]. Mortality rates are lower than those for cardioembolic stroke but higher than those for small artery disease.
Overall, the short-term risk of recurrent stroke after cryptogenic stroke is intermediate between the high early risk after large artery atherosclerosis stroke and low risk after small artery disease stroke. In the Oxford meta-analysis of four large population-based studies, the risk of recurrent stroke after cryptogenic stroke was 1.6 percent at seven days, 4.2 percent at one month, and 5.6 percent at three months . In the NINDS Stroke Data Bank, 3 percent of patients with cryptogenic stroke had recurrent events at one month . In the NOMASS study at three months, the risk of recurrence for the cryptogenic group was 3.7 percent , slightly lower than those found in the Oxford meta-analysis .
At two years, recurrence risk ranges from 14 to 20 percent [22,65,69]. In the Stroke Data Bank, cryptogenic stroke had the lowest two-year recurrence risk and was an independent predictor of low recurrence risk . At five years, the long-term recurrence risk was 33.2 percent in Rochester, not significantly different from the other subtypes .
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Stroke in adults".)
SUMMARY AND RECOMMENDATIONS
●Cryptogenic stroke is defined as brain infarction that is not attributable to a source of definite cardioembolism, large artery atherosclerosis, or small artery disease despite a thorough vascular, cardiac, and serologic evaluation. Embolic stroke of undetermined source (ESUS) is defined as a nonlacunar brain infarct without proximal arterial stenosis or cardioembolic sources. ESUS represents a subset of cryptogenic stroke. (See 'Classification' above.)
●The pathophysiology of cryptogenic stroke is likely heterogeneous. Proposed mechanisms include cardiac embolism secondary to occult paroxysmal atrial fibrillation, aortic atheromatous disease or other cardiac sources, paradoxical embolism from atrial septal abnormalities such as patent foramen ovale, hypercoagulable states, and preclinical or subclinical cerebrovascular disease. (See 'Possible mechanisms' above.)
●Cryptogenic stroke accounts for 25 to 40 percent of ischemic stroke. (See 'Epidemiology and risk factors' above.)
●Cryptogenic stroke presents with superficial hemispheric infarction in the majority of patients, and a significant proportion of cryptogenic strokes adhere to embolic infarct topography on brain imaging. (See 'Clinical features' above.)
●Cryptogenic stroke is a diagnosis of exclusion. The diagnosis is made when a standard evaluation reveals no definite evidence of cardioembolism, large artery atherosclerosis, small artery disease, or other determined etiology, and no evidence of atrial fibrillation on a 12-lead electrocardiogram (ECG) and on 24-hour cardiac monitoring. Additional studies can be pursued if the standard evaluation fails to determine the probable cause. We suggest ambulatory cardiac monitoring for several weeks (eg, 30 days) for all adult patients with a cryptogenic ischemic stroke or cryptogenic transient ischemic stroke (TIA). (See 'Evaluation and diagnosis' above.)
●The acute management of cryptogenic stroke is similar to that of other ischemic stroke subtypes. For secondary prevention, most patients with an ischemic stroke or TIA should be treated with blood pressure reduction, antithrombotic therapy, statin therapy, and lifestyle modification. However, the optimal antithrombotic therapy of patients with cryptogenic stroke who have atrial septal aneurysm, atheromatous aortic disease, or coagulation disorders is uncertain. (See 'Treatment' above.)
•For patients with a first cryptogenic stroke, we recommend antiplatelet therapy rather than anticoagulant therapy while awaiting the results of long-term cardiac monitoring (Grade 1B). (See 'Pending long-term cardiac monitoring' above.)
•For patients initially diagnosed with cryptogenic stroke who have atrial fibrillation of any duration detected on long-term monitoring, even if detected remotely from the incident stroke, we suggest anticoagulant therapy with warfarin or a non-vitamin K dependent oral anticoagulant (NOAC) rather than antiplatelet therapy (Grade 2C). (See 'Occult or subclinical atrial fibrillation' above.)
•Percutaneous PFO closure in addition to antiplatelet therapy is suggested for patients age ≤60 years with an embolic-appearing cryptogenic ischemic stroke who have a PFO with a right-to-left shunt detected by bubble study. (See 'Presence of a PFO' above.)
•For patients on antiplatelet therapy who have a recurrent cryptogenic stroke and no atrial fibrillation on re-evaluation with long-term cardiac monitoring, options include continuing the same antiplatelet agent or switching to another antiplatelet agent; for patients with recurrent ESUS, switching to empiric anticoagulant therapy is also a reasonable option. (See 'Recurrent cryptogenic stroke' above.)
●Compared with other stroke subtypes, cryptogenic stroke tends to have a better prognosis and lower long-term risk of recurrence. (See 'Prognosis' above.)