Overview

Angiography was first developed in 1927. The first cannulation of the right atrium occurred in 1929. In 1960, Mason Sones performed the first coronary angiogram (13). Over the ensuing decades, cardiac catheterization has represented a major tool in the armamentarium of both diagnostic and interventional cardiology. Cardiac catheterization, successfully performed over 1.5 million times annually in the United States, is not without risk.

Neurologic complications range from minor events that may or may not lead to any permanent sequelae (such as transient compression of the femoral nerve by subcutaneous hematoma) to catastrophic events (cerebrovascular arterial occlusion or dissection). The major focus of this article is to describe cerebrovascular complications of diagnostic and cardiac catheterization. We also review some of the more common peripheral nervous system complications that are neurologically related. Placing cardiac catheterization in a broader clinical perspective, the reader is directed to an excellent review of the historical evolution of diagnosis and treatment of coronary artery disease and myocardial infarction (72).

Neurologic complications can be divided into central nervous system and peripheral nervous system complications (Table 1). CNS complications of cardiac catheterization include cerebrovascular (eg, transient ischemic attack, ischemic stroke, intracerebral hemorrhage), infectious events (eg, embolization of infected material to the brain), retinopathy, and myelopathy. In addition, contrast agent use may provoke seizures. PNS complications include mononeuropathy (eg, femoral nerve compression), plexopathy (eg, due to retroperitoneal hematoma formation), and radiculopathy.

Major non-neurologic complications of cardiac catheterization include myocardial infarction perforation of the heart or great vessels, and death. Local arterial puncture-related complications include local hematoma formation, retroperitoneal hemorrhage, arteriovenous fistulas, pseudoaneurysm formation, and arterial thrombosis. Cardiac arrhythmias, atheroembolism, acute renal failure, infection, and radiation-related adverse effects may also occur.

Table 1. Neurologic Complications of Cardiac Catheterization

(Adapted from the article, Cardiovascular intervention: neurologic complications.)

Stroke

Stroke with cardiac catheterization may occur during the procedure, immediately following the procedure while the femoral artery sheath is still intact, or up to 36 hours following the procedure (89; 86). Strokes during diagnostic and therapeutic cardiac catheterizations (percutaneous coronary interventions) occur infrequently (Table 2). The mechanisms for many of these events are well understood (Table 3). However, unlike the spontaneously occurring stroke, the literature regarding treatment of stroke that occurs in the context of cardiac catheterization has been largely described in retrospective reports with relatively small patient sample sizes. The current understanding and limitations in treatment of strokes occurring in this circumstance are described below. Complications of other cardiac catheter-delivered therapies such as patent foramen ovale closure, transcatheter aortic valvular replacement, and atrial fibrillation ablation procedures are beyond the scope of this article but have been addressed elsewhere (12).

Table 2. Stroke Incidence: Registry Data Over Time

Table 3. Factors Influencing Stroke Following Percutaneous Coronary Intervention

(33; 62; 106; 35; 101; 53; 79; 98)

Stroke location and prognosis. In a single center retrospective review of 19,163 patients who underwent percutaneous coronary intervention over a 15-year time span, 89 patients experienced stroke (ischemic or hemorrhagic) or transient ischemic attack. Characteristics of the patients experiencing ischemic strokes included older age (73 vs. 66 years), female gender, history of hypertension, peripheral vascular disease, or myocardial infarction 7 days prior to percutaneous coronary intervention as well as intracoronary thrombus, cardiogenic shock proceeding the intervention, or emergent coronary intervention.

The majority of strokes (91%) were consistent with arterial embolization (peripheral cortical or gray-white junction location), with the remaining representing small subcortical infarctions. Although hypotension increased the likelihood of poor outcome, surprisingly those patients did not experience watershed infarctions. Eighty percent of the infarctions occurred in the middle cerebral artery, either singly or in conjunction with infarctions in other locations. Strokes involving the entirety of the middle cerebral artery distribution were uniformly fatal (median 6 days, range 1 to 10 days). Infarctions in a single vascular territory generally had good functional outcome with little or no residual disability at 3 and 6 months (modified Rankin scale 0-1), but involvement of multiple vascular territories produced significant neurologic deficit (modified Rankin scale greater than 3) (38).

Cerebrovascular events following percutaneous coronary interventions (both diagnostic and therapeutic) were associated with high rates of in-hospital complications, including a 3-fold risk of acute renal failure, a 4-fold risk of requiring dialysis, and an 8-fold adjusted risk of death (56; 28).

A proposed etiology of renal failure in this context is emboli to the kidneys. Similarly, according to one series, 1 in 5 patients with nonfatal cardioembolic stroke or transient ischemic attack from atrial fibrillation, the event may have been associated with infra-diaphragmatic visceral infarction of kidneys, spleen, liver, or bowel (90). Chronic renal insufficiency itself represents a potential underlying risk for stroke, which may cause platelet dysfunction or bleeding diathesis (28). A more detailed discussion of renal failure and stroke is presented below. Embolization to the central or branch retinal arteries during cardiac catheterization has a reported incidence up to 6.3% (52).

The angiographer should be vigilant for any visual symptoms that might require emergent attention in an effort to mitigate visual loss (52).

Stroke-related mortality has been reported as high 42%. In some series, almost half have been reported to be hemorrhagic, in part possibly due to intraprocedural antithrombotic therapy (33; 28; 03). Studies performed more recently have demonstrated significantly less hemorrhagic strokes, likely due to advances in intraprocedural antithrombotic therapy. Causes and circumstances of in-hospital mortality amongst patients undergoing percutaneous coronary intervention were reviewed in a single center retrospective study of 5520 patients treated over the first decade of this century. Of the 85 deaths (1.5%), the most common complications were left ventricular failure, neurologic compromise, and arrhythmia (35.3%, 16.5%, and 12.1% respectively). Only 6 patients (7%) were found to have preventable deaths (96).

Procedural aspects and embolic potential.

Aortic atheroma. The prevalence of aortic atherosclerosis has varied among studies of different populations, and the aortic arch atherosclerosis has been reported as high 62.2% (68; 84). Transesophageal echocardiography has demonstrated that complex plaque (ulcerated or mobile plaque) rather than plaque thickness alone (less than 4 mm vs. more than 4 mm) is associated with increased embolic stroke risk in the elderly, which is consistent across multiple ethnic groups (27). In addition to spontaneous embolization, aortic plaque is susceptible to mechanical disruption during cardiac catheterization procedures. Catheterization performed under emergent circumstances may increase the risk of contacting the vessel wall rather than maintaining the catheter closer to the center of the aortic lumen. Larger (greater than 6 French) and less flexible catheters have a greater likelihood of dislodging atheroma and producing major adverse events, including stroke (06).

Crossing the aortic valve. There is increased embolic risk during retrograde catheterization across the aortic valve. In a prospective study, MRI and clinical examinations within 48 hours before and after transaortic valvular catheterization demonstrated that 22% of 101 patients had a focal embolic pattern of restricted diffusion, with 3 patients (3%) exhibiting clinically abnormal examination findings (75). A control group without transaortic valvular catheter passage demonstrated no radiographic evidence of embolization.

Catheter tip and guidewire atheroma. Fibrin formation at the catheter tip and along the guidewire can represent a source of embolization. Periodic catheter flushing, especially during catheter changes, can mitigate this phenomenon. Additionally, aortic atheroma can adhere to the guidewire and sheath, representing a source of embolization.

Unexpected "other" etiologies include previously undetected carotid artery steno-occlusive disease and intracranial hemorrhage. Dukkipati and colleagues reported that 4 of 92 patients in their series were discovered to have carotid stenosis ipsilateral to their symptoms (28). Furthermore, up to 50% of strokes following cardiac catheterization were hemorrhagic (33). These studies should serve as a reminder that not all strokes occurring in the context of cardiac catheterization are embolic.

Are emboli and diffusion-weighted MRI abnormalities clinically evident? Diffusion-weighted MRI is a commonly used and highly sensitive MR sequence to the early changes in acute ischemic stroke. This technique measures the random motion of water molecules, which is changed shortly after ischemic cell death. Increased DWI signal is seen within a few minutes after the vascular event. On another note, early diffusion-weighted MRI findings in ischemic stroke can be reversed, especially in the setting of early reperfusion. However, most frequently this reversal is transient, although true reversal may also occur.

Studies with no clinical symptoms. A study of the incidence of clinically silent infarctions following cardiac catheterization prospectively evaluated 52 patients with negative pre-catheterization MR imaging and found that 7 of 48 (15%) experienced one or multiple areas of focal restricted diffusion less than 5 mm in an embolic pattern distribution, without evidence of watershed type infarctions (17). All patients demonstrated unremarkable clinical neurologic examinations. Catheterizations were virtually evenly divided between diagnostic and therapeutic coronary interventions. The rate of asymptomatic cerebral infarction after catheterization was 10-fold greater than for clinically apparent cerebral lesions (0.11% to 0.38%). The only independent predictor of infarction was increased procedure time. Considerations included thrombus formation on the catheter tip, dislodging of atheroma from the ascending aorta, or air embolism (17).

Different types of emboli may occur at different stages of cardiac catheterization. Transcranial Doppler right middle cerebral artery signal analysis during diagnostic cardiac catheterization demonstrated that all 17 patients in a study had embolic signal during 2 major phases of the procedure (103). The first phase was during catheter and guidewire manipulation, with signal characteristics consistent with solid emboli. The second phase occurred during catheter flushing and contrast injection, consistent with microbubbles. Solid emboli may occur with mechanical catheter or injection jet disruption of atheroma or with catheter tip thrombus. Microbubbles may occur from agitated saline or rate of injection of contrast. Lund and colleagues reported that 92.1% of the microemboli were gaseous and 7.9% were solid (63). Postprocedural MRI demonstrated that infarctions occurred in 15.2% of patients during a transradial approach and none during transfemoral catheterization; however, this was not consistently described in other studies.

Studies with positive clinical symptoms. In a prospective study, MR imaging and clinical examinations within 48 hours before and after transaortic valvular catheterization demonstrated that 22% of 101 patients had focal embolic pattern restricted diffusion (75). Three patients (3%) had clinically evident examination abnormalities. A control group without transaortic valvular catheter passage demonstrated no radiographic evidence of embolization. Cognitive impairment as a sequela of catheter-induced microemboli has not been studied.

Overview of treatment. The low incidence of stroke during cardiac catheterization makes randomized, prospective trials difficult to perform. The 2018 Guidelines for the Early Management of Patients with Acute Ischemic Stroke recommend intravenous alteplase as a reasonable treatment of acute ischemic stroke complications of cardiac procedures depending on the usual eligibility criteria (Class of Recommendation IIa; Level of Evidence A) (80).

Published experiences detail retrospective treatment using intravenous and intra-arterial thrombolytic therapy. More recently, combined multimodal intra-arterial thrombolysis and endovascular clot disruption show generally favorable outcomes, albeit not always (89; 105; 48; 49; 03; 15). As the occurrence of stroke in the context of cardiac catheterization is infrequent, there is need for prospective randomized multicenter evaluation to obtain an adequate sample size to offer meaningful evaluation and treatment recommendations (see proposed management strategies).

The following represents specific broad categories that have been previously addressed in the literature, as well as those that need to be addressed in the future, in an effort to move towards a unified evaluation and treatment strategy.

If time is brain, should we add additional time to image the brain before treatment? Some investigators have advocated immediate cerebral angiography for intra-arterial thrombolysis while the coronary catheter or sheath is in place, without cerebral CT or MRI imaging, in an effort to minimize the time between onset of symptoms to attempted recanalization (24; 34). But this approach is risky and requires a reasonable and sufficient degree of certainty that ischemic stroke is the exact cause of present symptoms as the incidence of intracerebral hemorrhage following cardiac catheterization can be as high as 50% (64). Although more recent studies have demonstrated a lower incidence of intracerebral hemorrhage, there is uncertainty in the estimates of the benefit (53; 38). Another excellent option for neurovascular imaging is the cone-beam CT angiography (CBCT-A), which is surely available in all modern angiography suites and is quick to use. This fairly promising approach appears to shorten the time between stroke onsite and revascularization treatment (73).

In conclusion, there remains no safe way to exclude intracerebral hemorrhage without a rapid CT or MRI before exposing the patient to thrombolytic therapy, and the standard emergency evaluation currently includes brain and vascular imaging with CT or MRI.

"It makes sense to set up a Code Stroke protocol in every cath lab." In an editorial by Lyden, this recommendation was based on available evidence that patients experienced improvement following thrombolytic treatment. He further advocated in favor of expedited cerebral imaging to exclude the possibility of intracerebral hemorrhage, which can represent up to 50% of strokes occurring in this context (33; 24; 64). However, given the infrequent occurrence of strokes during cardiac catheterization and technical advances in percutaneous coronary interventions, any cardiac catheterization-related stroke treatment plan should be continually reviewed and updated to guide therapy based on those changes. The following represents a discussion of some of the salient issues.

The evolving use and combination of fibrinolysis, platelet anti-aggregants, and thrombin inhibitors need to be considered when developing a treatment strategy of cardiac catheterization-associated acute stroke to minimize potential complications. It is well known that systemic and intracoronary thrombolytic therapy for treatment of the acute myocardial infarction is a risk factor for both hemorrhagic and nonhemorrhagic strokes (100; 22; 65). It is worth mentioning that the diagnostic cardiac catheterization technique has minimized and often eliminated heparin as a method of clearing the catheter tip clots. Furthermore, therapeutic coronary interventions have shifted from intracoronary thrombolysis alone to angioplasty and stent placement. On another note, the use of unfractionated heparin in percutaneous coronary interventions has evolved to the use of GP IIb/IIIa antagonist plus reduced-dosage heparin. More recently, direct thrombin inhibitors (ie, bivalirudin) have demonstrated both non-inferiority and a 40% relative risk reduction in major bleeding complications (of which a sizable component occurred within the CNS), in comparison to the combined use of unfractionated heparin and GP IIb/IIIa antagonists (95). Bivalirudin has the advantage of avoiding heparin-induced thrombocytopenia and heparin-induced thrombosis syndrome and has a short half-life (25 minutes) in the context of normal renal function (99; 70). It can be used alone or if needed, in conjunction with a GP IIb/IIIa inhibitor.

Intravenous versus intra-arterial thrombolytic treatment strategies. A limited number of studies have been performed utilizing Intravenous versus intra-arterial thrombolysis for acute stroke following cardiac catheterization. Thus far, the published sample size is small, demonstrating a need for additional prospective randomized study. Both routes have demonstrated clinical benefit in small nonrandomized trials. Currently, diagnostic cardiac catheterization performed in many centers uses only minimal or no anticoagulant. However, in some centers, percutaneous coronary interventions (angioplasty, stent placement, etc.) for the patient experiencing ST-segment myocardial infarction or unstable angina may still use unfractionated heparin, often in conjunction with a GP IIb/IIIa inhibitor. The combination is associated with increased risk of major bleeding complications (02; 16; 88). Further study is needed to determine which stroke treatment modality is the safer alternative, or in the case of embolization to multiple vascular territories that would preclude intra-arterial therapy, the relative risk of systemic thrombolytic administration. Also deserving of further study, mechanical clot retrieval of a proximal arterial occlusion without thrombolytic therapy may be beneficial when concurrent anticoagulant therapy poses increased risk of intracranial hemorrhage or if there is radiographic demonstration of a calcific embolus (50; 19).

Safety of combined glycoprotein IIb/IIIa agents, anticoagulant, and fibrinolytic therapy. In 2001, analysis of pooled multi-trial data reported that the GP IIb/IIIa inhibitor abciximab, when used alone, demonstrated no significant difference in hemorrhagic and nonhemorrhagic strokes. However, there was an increased risk of hemorrhagic stroke when used in conjunction with standard-dose heparin, which could be overcome with heparin dosage reduction (02). In the GUSTO V trial, patients undergoing treatment for acute myocardial infarction were randomized to either receive standard-dose reteplase or half-dose reteplase and full-dose abciximab. Both groups received unfractionated heparin infusion in conjunction with and according to their assigned treatment strategies. No differences in overall rates of intracranial hemorrhage were observed between the two groups (88). However, older patients had a higher risk of hemorrhage, with younger patients having a lower hemorrhagic risk. Another analysis of 450 adverse events reported to the FDA Medwatch program, consisting of patient deaths following GP IIb/IIIa (eptifibatide-23%, tirofiban-32%, and abciximab-46%) use, showed that hemorrhage was a significant risk factor, with CNS the most common site of bleeding (greater than 35%) (16). Even more importantly, the hemorrhages occurred in the context of combined therapy with heparin (68%), aspirin (39%), clopidogrel or ticlopidine (25%), and thrombolytics (8%). The 2018 Guidelines for the Early Management of Patients with Acute Ischemic Stroke recommend against the use of abciximab concurrently with intravenous alteplase (Class of Recommendation III: Harm; Level of Evidence: B-R) (80).

Coronary artery stents, anticoagulant and antiplatelet therapy, and bleeding risks. Newly placed drug-eluting and bare-metal coronary artery stents require initial dual antiplatelet therapy (DAPT) using aspirin plus clopidogrel, prasugrel or ticagrelor to maintain patency. DAPT following coronary artery stent placement increases the risk for hemorrhagic complications, including intracranial hemorrhage (10). One element in the decision to use of a drug-eluting stent versus a bare-metal stent is the potential risk for in-stent thrombosis. Drug-eluting stents by design are slower to endothelialize and require a longer duration of antiplatelet agents than bare-metal stents. Although the rate of stent thrombosis was higher in bare-metal stents, data from the largest registry to date did not demonstrate any significant difference in rates of all-cause mortality, nonfatal myocardial infarction, stroke, transient ischemic attack, major and minor bleeding events, and total adverse events (51).

The presence of atrial fibrillation, mural cardiac thrombus, mechanical cardiac valves, etc. often necessitate therapeutic anticoagulation to minimize the risk of cardiogenic embolization. Therapeutic anticoagulation alone is insufficient to prevent coronary stent thrombosis, which requires initial DAPT. The Active-W trial has demonstrated superiority of warfarin over antiplatelet therapy in prevention of stroke from atrial fibrillation (21). Coronary stent placement in the presence of atrial fibrillation generally necessitates both DAPT and anticoagulant therapy, creating a 3-fold risk in hospital admission for bleeding complications compared to solitary oral anticoagulant therapy (91). In this context, triple therapy using DAPT plus therapeutic anticoagulation reduces the risk of major adverse clinical events (stroke, stent thrombosis, myocardial infarction, systemic embolization, and all-cause mortality) but also increases the risk for hemorrhagic complications (55; 83).

Duration of single versus DAPT is predicated on coronary stent type (ie, bare-metal versus drug-eluting stent), the circumstances of stent placement (acute coronary syndrome versus elective percutaneous coronary intervention) and coexisting conditions that may also predispose to hemorrhage. Drug-eluting stents require a longer time to endothelialize, necessitating a longer duration of dual antiplatelet therapy than bare-metal stents. Premature discontinuation of DAPT in patients with drug-eluting stents is associated with stent thrombosis and death, leading to the recommendation of DAPT for 12 months in the absence of increased bleeding risk (40; 39; 93; 59).

Although a major practice guideline has not supported the DAPT longer than 1 year based on the high risk-benefit ratio (97), the duration of DAPT administration has not been firmly established and remains a subject of continued study. A meta-analysis of 4 randomized controlled trials comparing DAPT of 3 to 6 months duration in comparison to 12 to 24 months revealed a statistically significant increase in hemorrhagic events in the longer duration group, with a nonstatistically significant decrease in stent thrombosis (29). A randomized controlled trial comparing 12 versus 30 months of continued DAPT versus aspirin plus placebo after 1 year of dual antiplatelet therapy in patients with drug-eluting stents demonstrated reduced rates of stent thrombosis (0.4 versus 1.4%; P < 0.001), myocardial infarction, and stroke (4.3% versus 5.9%; P < 0.001) in the extended DAPT group counterbalanced against an almost equal increased risk of moderate or severe hemorrhagic complications (2.5% versus 1.6%; P = 0.001) (66). Prolonged DAPT after 1 year needed to maintain drug-eluting stent patency might have a possible, incremental benefit in yet-to-be defined subsets of patients (20). To enhance the safety profile, evaluation of short duration clopidogrel (6 weeks vs 6 months) component of triple drug therapy in patients with drug-eluting stents is currently being investigated (31). The WOEST study compared oral anticoagulant therapy plus clopidogrel with triple drug therapy, demonstrating a substantial reduction in bleeding events (19.4% versus 44.4%, p < 0.0001) without significant increase in thrombotic events. There was no difference in the rate of intracranial hemorrhage between groups (1.1%) (26). Novel oral anticoagulants and alternate antiplatelet agents have also being studied (04; 104).

Catheter access site and related complications. Cardiac catheterization may be achieved via access from the femoral or radial arteries. Table 3 highlights major peripheral nerve injuries related to cardiac catheterization. Because the majority of cardiac catheterization procedures are performed via transfemoral route, Table 4 provides a more detailed summary of the most common arterial access-related complications, their pathophysiology, and potential neurologic complications. Local complications of femoral artery puncture may lead to direct trauma of the adjacent femoral nerve or from extrinsic compression of the femoral nerve or lumbosacral plexus by hematoma formation in the upper thigh or retroperitoneum respectively. Nerves distal to the site of hematoma formation may be compromised by ischemic monomelic neuropathy (Souza et al 1998; 61; 11).

Table 4. Peripheral Nervous System Complications of Cardiac Catheterization

Adapted from (37).

Table 5. Complications of Femoral Arterial Cannulation

Radial artery catheterization was introduced in 1989. Meta-analysis has demonstrated a 2-fold reduction in the onset of death and a 1.5-fold reduction in major adverse clinical events in ST segment myocardial infarction patients undergoing percutaneous coronary intervention via transradial approach. The superficial position of the radial artery allows for early recognition of bleeding complications and easy and rapid compressibility. Another meta-analysis of randomized controlled trials enrolling a total of 22,843 patients with coronary artery disease undergoing percutaneous coronary intervention has proved the safety of the transradial approach showing significant reductions in mortality rates and major adverse cardiovascular events as well as major bleeding and vascular complications with radial access (30). A recent comparative prospective study conducted among 206 patients with acute myocardial infarction has shown an advantage of emergency percutaneous coronary intervention via transradial approach, leading to less vascular complications compared with femoral access (81). Stroke risk was similar to the transfemoral approach (43; 78). The smaller diameter of the radial artery in comparison to the femoral artery may increase the risk for vascular avulsion from a relatively larger diameter catheter sheath (69). When radial artery access cannot be achieved due to occlusion, spasm, or radial loops that are difficult to navigate, ulnar artery catheterization has demonstrated efficacy and low incidence of adverse sequela. Three studies reporting a total of 564 patients demonstrated only 1 ulnar nerve injury, 3 major hematomas requiring intervention, 8 minor hematomas, and no cases of limb ischemia (54; 25; 44).

Local hematoma formation. A large area of thigh ecchymosis around the femoral artery puncture site is not rare. Subcutaneous hematoma may compress the femoral nerve. Femoral nerve compression may last weeks to months to resolve. Most patients only have sensory discomfort, including paresthesias, dysesthesias, and pain. Decreased strength of femoral nerve-innervated muscles may occur. Avoiding strenuous activity may prevent expansion of hematoma formation and, hence, femoral nerve compression.

Retroperitoneal hematoma. Femoral artery puncture above the inguinal ligament may result in hematoma extension into the retroperitoneal space. The retroperitoneum is divided into the perirenal space and anterior and posterior pararenal spaces. Direct hemorrhage within the psoas muscle may also occur. The lumbar plexus and its branches lay anterior to the psoas muscles. A retroperitoneal hematoma may compress the lumbar plexus, particularly the lower lumbar plexus, the sacral plexus, and the femoral or obturator nerves (14). Fascial tracking of blood in the femoral nerve sheath may result in femoral neuropathy in the absence of visible retroperitoneal blood.

Kent and colleagues reported a 0.5% incidence of retroperitoneal hematoma in a retrospective review of 9200 femoral artery catheterizations (47). Clinical manifestations of retroperitoneal hematoma include suprainguinal tenderness, low back pain, lower quadrant pain, and thigh paresthesias or dysesthesias. Thigh adduction weakness may suggest compression of the obturator nerve. CT scan of the abdomen is the diagnostic test of choice to establish the diagnosis. Treatment is usually conservative. Some patients may require blood transfusions. Surgical intervention is rarely necessary. The outcome of lumbosacral plexopathy secondary to retroperitoneal hematoma is generally favorable. In a retrospective study, 5 of 16 patients complained of mild sensory symptoms at long-term follow-up (46).

Arteriovenous fistula. Concomitant puncture of the femoral vein may lead to a fistulous communication between the femoral artery and the femoral vein. A thrill or continuous bruit at the puncture site is the most common manifestation. Ischemic monomelic neuropathy may occur following arterial stenosis or occlusion that produces limb ischemia with coexisting multiple distal mononeuropathies distal to the site of occlusion. Risk factors include peripheral vascular disease, abdominal aortic aneurysm, low cardiac output, and intra-aortic balloon pump use (07). Arterial ultrasound or arteriography may confirm the diagnosis, which may be treated surgically.

Pseudoaneurysm formation. Femoral pseudoaneurysm occurs when a periarterial hematoma remains in continuity with the arterial lumen. A pulsatile mass with a systolic bruit within 3 days of femoral arterial puncture is a common presentation of pseudoaneurysm formation. Doppler ultrasound may establish the diagnosis. Treatment options include surgical repair, ultrasound-guided compression, or ultrasound-guided local injection of thrombin or collagen. Recently, Babunashvili and colleagues have proposed a new technique of post-catheterization radial artery pseudoaneurysm treatment by placing a long arterial sheath directly into the radial artery, applying gentle mechanical compression, or injecting thrombin to isolate the pseudoaneurysm sac with subsequent sac thrombosis and closure (08). The incidence of femoral neuropathy was 2.5% in a prospective study of 79 patients treated for femoral pseudoaneurysm following cardiac catheterization (85). In addition, femoral mononeuropathy has been reported in traumatic pseudoaneurysm formation associated with iliacus muscle hematoma (82).

Effects of renal disease. Iodine-based radiographic contrast agent for cardiac catheterization is a well-established cause of acute kidney injury. Of additional importance is the relationship of renal failure in stroke type, severity, hemorrhagic conversion, and treatment.

Diabetics are at greater risk of contrast-induced acute kidney injury. The risk is further elevated in patients with hypercholesterolemia and low left-ventricular ejection fractions lower than 40% (77). Elevated pre-procedural glucose in nondiabetic patients is associated with a greater risk of contrast-induced acute kidney injury (94). Rosuvastatin therapy in statin-naïve acute coronary syndrome has been demonstrated to significantly reduce the risk of acute kidney injury as well as the 30-day incidence of death, dialysis, myocardial infarction, stroke, and persistent renal injury (3.6% vs. 7.9%) (36; 58). The proposed mechanisms of statin protection include its antiinflammatory properties independent of its cholesterol-lowering effect; improvement of endothelial function by increasing nitric oxide synthetase bioavailability, which decreases oxidative stress; and its anti-apoptotic effects (05).

Chronic kidney disease is highly prevalent among hospitalized patients with stroke. This patient population has a higher prevalence of hemorrhagic transformation, cerebral microbleeds, and systemic hemorrhages (42). Chronic kidney disease has a higher incidence of non-risk-adjusted symptomatic intracranial hemorrhage and serious systemic hemorrhage in comparison to patients with normal renal function. However, when adjusted for risk factors, there was no clear trend of increased incidence of hemorrhage with worsening stages of renal dysfunction. However, patients with chronic kidney disease had a substantial (23%) increase in the overall incidence of poor outcome (in-hospital mortality and lack of independent ambulation at discharge) with worsening renal function. The potential for hemorrhage in the presence of chronic kidney disease should not represent a contraindication to intravenous tPA (76).

Cardiac arrhythmias. Cardiac arrhythmias complicating cardiac catheterization include ventricular premature beats, ventricular tachycardia or fibrillation (0.2%), atrial arrhythmias, bradycardia, and conduction disturbances (41). Atrial fibrillation may lead to cardioembolism to the brain or spinal cord. Both tachyarrhythmias and bradyarrhythmias may produce hypotension and potential for ischemic brain injury.

Heart and great vessels perforation. Procedures with stiffer catheters are at higher risk of perforating the heart or great vessels. These procedures include endomyocardial biopsy, balloon valvuloplasty, transseptal catheterization, pericardiocentesis, and placement of pacing catheters. Neurologic sequelae in patients who survive these complications include hypoxic ischemic brain injury or ischemic myelopathy.

Atheroembolism. Catheter advancement through arterial lumen may lead to mechanical disruption of atheroma. The resultant debris may embolize to the retina, brain, or spinal cord. Similarly, ischemic monomelic neuropathy may occur following embolization to limb vasculature. A prospective study of 1000 percutaneous cardiac intervention patients showed that in more than 50% of interventions, guiding catheter placement was associated with scraping atherosclerotic debris from the aorta (45). Cholesterol embolization syndrome (CES) may produce livedo reticularis, blue toe syndrome, digital gangrene, eosinophilia, and renal dysfunction. The incidence of cauda equine syndrome was 1.4% in a prospective study of 1786 patients who underwent cardiac catheterization (09).

Infection. The femoral arterial approach has a lower risk of infection (0.06%) compared to the brachial artery approach (0.6%) (18). Infrequently, if endocarditis ensues, a cerebral mycotic aneurysm may form.

In 2017, Means and colleagues reviewed available data on the management of common vascular complications occurring during and after percutaneous coronary artery interventions. These complications include pseudoaneurysms, retroperitoneal bleeding, arteriovenous fistulae, radial artery occlusion, stent thrombosis, incomplete stent apposition, stent fracture, stent underexpansion, dissection of coronary arteries, perforation of coronary arteries, and pericardial effusion or tamponade as well as decreased or absence of flow in large epicardial arteries (67).

Myocardial infarction. Risk of myocardial infarction during cardiac catheterization is approximately 0.1% (102). Myocardial infarction has a substantial increased risk of stroke. Analysis of 111,023 discharged patients with myocardial infarction between 1994 and 1995 revealed that independent predictors of ischemic stroke following myocardial infarction included age greater than 75 years, black race, no aspirin at discharge, atrial fibrillation, diabetes, hypertension, prior history of stroke, and history of peripheral vascular disease. Following discharge diagnosis of myocardial infarction, 2.5% of patients had ischemic stroke at 6-month follow-up. It is not known if myocardial infarction following cardiac catheterization poses an additional increased risk of stroke in the long-term (60).

Contrast- and medication-related complications. Major contrast-related complications include nephropathy and seizures. History of diabetes mellitus and chronic kidney disease increases the risk of contrast nephropathy. Renal failure may raise blood pressure with the potential for ischemic stroke or intracerebral hemorrhage. Renal insufficiency will be discussed in more detail in the next section.

Seizures following cardiac catheterization may be due to stroke or contrast agents. Overall, contrast-induced seizures have mostly been reported in adults but may also occur in children. Contrast-induced seizures were reported in a 6-year-old child undergoing cardiac catheterization (87). Partial seizures from contrast agent retention were reported in an 18-month-old infant who underwent cardiac catheterization (32). Although the risks of various iodinated contrast reactions have been reduced with iso-osmolar nonionic contrast agents, they still exist, especially when a large cumulative dose of contrast has been used (32; 57).

Proposed management strategies

The clinical management of patients who suffer cerebrovascular complications during or following cardiac catheterization is unlike that applicable to any other clinical scenario. More often than not, the patients have received sedatives that obscured their neurologic assessment and are likely to have received antithrombotic agents that notably increase the risk of hemorrhagic stroke but limit the applicability of intravenous thrombolytic treatment.

In general, the approach to patients who suffer an acute neurologic change while still in the catheterization suite is materially different than that of patients whose neurologic symptoms are discovered after the cardiac procedure has been completed. However, they do share some common steps during the early treatment stages:

Once these two steps are completed, the subsequent strategy will depend largely on the circumstances of the discovery of the acute neurologic symptoms.

Acute stroke "on-the-table."

On recognition of acute neurologic changes during a heart procedure, and following the two common steps described above, if the patient is not back at his baseline following the administration of sedation reversing drugs, the next step depends on the availability in that particular catheterization suite of cone-beam CT. This modality, increasingly available in modern catheterization suites, allows rapid, "on the spot" image acquisition of the brain without having to move the patient. If this is carried out and a hemorrhagic stroke is diagnosed, the cardiac index procedure should be aborted and the patient moved to the neurointensive care unit for further care. If, on the other hand, there is no imaging evidence of an intracranial hemorrhage, urgent neuroangiography should be the next step because it can be carried out without moving the patient.

Although in many hospitals cardiologists have been credentialed to perform "carotid angiograms," we must emphasize the need for an expert neuroangiographer to take over the care of the patient at this moment. A lot of the findings in the context of this clinical scenario may be sufficiently subtle to require experience that cardiologists do not generally have. In fact, a close look at the figure above shows that, if cone-beam CT is not available, we recommend proceeding with urgent neuroangiography anyway because it is likely to uncover the most important information relevant to the immediate care of the patient.

If a large arterial occlusion is diagnosed during the neuroangiographic study, the operator can then quickly proceed to thrombectomy. Moreover, irrespective of the presence of a large artery occlusion, at that moment the patient can be qualified for treatment with intervenous tPA.

Acute stroke "off-the-table."

Unlike the patients described in the previous section, the recognition of acute neurologic changes typically occurs after the cardiac index procedure has been concluded and the patient is in the recovery area or the coronary care unit shortly thereafter. Again, following the 2 common steps described earlier, if the patient is not back at his baseline following the administration of sedation-reversing drugs, the next step is to acquire urgent imaging of the brain and its vasculature by means of non-contrast CT and CTA studies. In addition, and although still controversial, some recommend adding perfusion CT studies to the urgent imaging protocol. In this context, 2 caveats are important to point out:

Once again, if the urgent CT imaging demonstrates a hemorrhagic stroke, the patient should be immediately transferred to the neurointensive care unit for further care per clinical practice guidelines. Conversely, if the CTA demonstrates a large arterial occlusion, the patient should be immediately transferred back to the catheterization suite for urgent thrombectomy. In parallel, the patient can be qualified for treatment with intravenous tPA.

Intravenous thrombolysis qualification and administration. In general, qualifying and administering intravenous tPA to one of these patients is not materially different than in any other clinical scenario. However, there are some details that are worth mentioning, the first one regarding how to manage those patients who have received unfractionated heparin during the cardiac index procedure. Clearly, the general recommendations are that active anticoagulation (ie, aPTT > 40 sec) is a contraindication for treatment with intravenous tPA. However, there is an increasing number of reports in the literature of patients successfully treated with intravenous thrombolysis following reversal of the heparin-related anticoagulation. Therefore, we recommend prompt measurement of both activated clotting time and activated partial thromboplastin time in these patients, followed by intravenous administration of protamine sulfate in sufficient doses to normalize the coagulation parameters.

It is also important to note that imaging of the brain using non-contrast CT is a sine qua non criterion for treatment with intravenous thrombolysis. Therefore, those patients who have not had such test should promptly undergo imaging, including the same modalities described above and with the same caveats. Once this is accomplished and the heparin has been reversed, the next major hurdle has to do with the patient having received other antithrombotic agents, namely glycoprotein IIb/.IIIa inhibitors or bivalirudin. Presently, any of these agents represents a substantial increased risk for hemorrhagic complications following the administration of intravenous tPA and, therefore, one should probably avoid such treatment.

Otherwise, it should be possible to qualify the patient's benefits-to-risks profile using a standard approach for the use of intravenous tPA in patients with acute ischemic stroke. Those patients who otherwise qualify should be promptly treated with intravenous tPA, and the attention of the treatment team should then also include management of the arteriotomy site.

Arteriotomy management protocol. In our opinion, having an arterial access site for the cardiac index procedure should not necessarily preclude the use of intravenous tPA for the treatment of these patients because they all represent "compressible" sites.

There is a slight difference in the management of femoral arteriotomies because of the size of the vessel as well as its potential for percutaneous closure. Along these lines, if the patient has a femoral sheath in place at the moment of the discovery of the acute neurologic changes, 2 options are available: (1) percutaneous closure, and (2) suturing the sheath with delayed removal. If the former path is chosen, we favor the use of closure devices to secure the arterial wall. Following closure, however, care of the arteriotomy site should be quite similar to that of a patient with a recently removed sheath due to the possibility of leakage following thrombolytic treatment.

Patients who have radial or brachial arteriotomies should always have their sheath removed because of the thrombotic complications associated with smaller arteries. Following removal of any arteriotomy sheath, pressure dressings must be available to control any thrombolytic-related bleeding. Femoral arteriotomies are best handled using a combination of femoral artery compression device with Doppler monitoring of distal pulses. The device should be inflated to a pressure sufficient to contain the bleeding, while producing minimal effect on the distal Doppler signal. Radial arteriotomies can be handled similarly by using the radial artery compression device; in most of these cases, Doppler monitoring is unnecessary due to the collateral circulation to the hand. Finally, the most threaded arteriotomies involve the brachial artery, which is an unforgiving vessel. No specific devices for bleeding control are available for this site; we recommend using a standard blood pressure cuff position proximal to the arteriotomy site and always in conjunction with Doppler monitoring of the distal circulation.

Additional comments. At the time of this writing, intravenous tPA almost invariably encompasses the use of intravenous alteplase. This is due to the fact that this is the only currently approved form of tPA for treatment of acute ischemic stroke. Thus, although some literature exists on the use of other similar drugs (eg, tenecteplase) the evidence for using these is insufficient for recommendation. That said, it is conceivable that in the near future more experience with different drugs would be available to improve our pharmacologic resources.

On another note, regarding the topic of large arterial occlusion, the bulk of the current prospective data refers to treatment of patients with terminal internal carotid artery or middle cerebral artery trunk occlusion. However, there is an increasing body of literature suggesting that thrombectomy may be comparably effective in patients with acute occlusion of the basilar artery, or even of more distal middle cerebral artery (eg, M2) arterial branches. Therefore, qualification and treatment of these patients using these techniques requires expert judgment and benefits-to-risk assessment.

Clinical vignette

Case 1. A 64-year-old woman with past medical history of hepatitis C-induced cirrhosis necessitating liver transplant, hyperlipidemia, and diabetes mellitus presented to the emergency department with chest pain caused by an ST-elevated myocardial infarction. Emergent cardiac catheterization demonstrated left main coronary artery disease treated with percutaneous coronary intervention. Following the procedure, she demonstrated slurred speech, left hemiparesis, right gaze preference, left hemineglect, and an extensor left plantar response. Initial cerebral CT was unremarkable. Laboratory studies were notable only for a chronically elevated white count of 26,500. Potential thrombolytic therapy would preclude a loading dosage of clopidogrel, placing the stent at risk for thrombosis. After a detailed discussion with the family and the cardiologist, intravenous rt-PA was administered, without appreciable clinical benefit. Subsequent CTA of the intracranial vasculature demonstrated truncation of an M2 branch of the right middle cerebral artery consistent with acute occlusion.

An incidental large cavitary right upper lobe pulmonary lesion was discovered. Fungal elements were cultured from a bronchioalveolar lavage. Needle biopsy of the pulmonary apical lesion was deferred in view of necessity for dual antiplatelet therapy to maintain stent patency over the short-term. She was treated with empiric voriconazole and followed with surveillance chest CT imaging. Brain MRI demonstrated extensive restricted diffusion in the right frontotemporal region with borderline increased T2/FLAIR hyperintensity compatible with acute or early subacute infarct.

Case 2. Following a prolonged right and left heart catheterization, we were consulted to evaluate a 42-year-old man with bilateral upper extremity (R > L) weakness. Pertinent past medical history included dextrocardia with situs solitus, so-called “congenitally corrected” transposition of the great arteries (eg, ventricular inversion with left-anterior aortic malposition), multiple ventricular septal defects, and valvular pulmonic stenosis. He had prior ischemic strokes with old areas of left frontal and right striatal encephalomalacia on CT. In addition, he had evidence of abnormal segmentation of C5 and C6, with a 5 mm posterior subluxation of C5 with respect to C6 on cervical spine x-rays.

Pertinent findings on neurologic examination included a “waiter’s tip” posturing of the upper extremities. There was marked weakness (R > L) of shoulder abduction and elbow flexion with the supinated arm, weak flexion of the forearm with the forearm midway between pronation and supination, weak external rotation of the upper arm, and preservation of shoulder elevation. Biceps and brachioradialis reflexes were absent. Triceps reflexes were preserved. There were bilateral Hoffmann responses. He had bilateral patellar hyperreflexia with positive suprapatellar reflexes and bilateral crossed adductor responses. Ankle reflexes were brisk without clonus. Both plantar responses were extensor. There was minimal decreased sensation to pinprick on the medial aspect of the left hand.

A clinical diagnosis of bilateral Erb type of brachial plexopathies with superimposed upper motor neuron findings from prior stroke was made. EMG showed acute denervation in C5, C6 innervated muscles bilaterally. Motor axons were anatomically continuous to some degree. The patient received intensive physical therapy and had a complete recovery of his bilateral brachial plexopathies at 6-month follow-up.