Thrombolysis of M1 Middle Cerebral Artery Occlusion

A 50-year-old man was admitted with a mild left-sided weakness which had developed 40 minutes before presentation. Initial neurologic examination revealed a left-sided pronator drift during the arm pronation test. No vascular risk factors were known. During emergency CT he clinically deteriorated with progression to a high-grade hemiparesis with dysarthria (National Institutes of Health Stroke Scale [NIHSS] score: 9).

Acute MCA ischemia caused by right distal M1-MCA occlusion of unknown origin.

After exclusion of contraindications, thrombolysis with 75 mg intravenous recombinant tissue plasminogen activator (IV rt-PA) was commenced 1.5 hours after the onset of symptoms.

• Was there evidence of atherosclerosis in the extracranial brain-supplying arteries?
  
• Could there be a sustained occlusion of the right MCA?
  
• If so, was there evidence of collateral blood flow via the anterior cerebral artery (ACA) and posterior cerebral artery (PCA)?

(Performed at the same time that thrombolysis was initiated.)

Extracranial Duplex Sonography

Color-coded imaging of the extracranial vessels showed no atherosclerotic vascular changes. Doppler spectrum analysis revealed no relevant differences in bilateral flow velocities or pulsatility (not shown).

Transcranial Duplex Sonography

During thrombolysis the neurologic status of the patient improved steadily.

Transcranial Duplex Sonography

Cardioembolic occlusion of the right distal M1-MCA segment with recanalization during intravenous thrombolysis with rt-PA.

Clinical Aspects

Here, we discuss a 50-year-old stroke patient who initially presented with a mild left-sided hemisyndrome, which progressed after 70 minutes to high-grade hemiparesis. He received IV rt-PA thrombolysis 90 minutes after symptom onset. The underlying cause was a cardiac embolism.

Our patient’s early clinical deterioration was rapidly assumed to be caused by clot disruption and distal fragmentation. Neurologic worsening in acute ischemic stroke is poorly defined. Analyzing 1964 consecutive patients admitted within 4 hours of onset, 256 patients (13%) had a worsening after 48–72 hours defined by an increase of at least 1 point on the NIHSS. Neurologic worsening was considered a progressive stroke in 33.6% of cases, increased intracranial pressure in 27.3%, recurrent ischemia in 11.3%, and secondary parenchymal hemorrhage in 10.5%. ICA or M1-MCA occlusion, territorial infarction, brainstem infarction, and diabetes mellitus were identified as independent predictors (Weimar et al 2005). In the earliest phase of acute stroke increased intracranial pressure can almost be ruled out, but in the case of rt-PA treatment urgent CT is required. If bleeding is excluded, a clot fragmentation and secondary distal vessel occlusion, progressive in-situ atherothrombotic occlusion of a main-stem vessel or orifice of a perforator artery, or a hemodynamically caused hypoperfusion by arterial hypotension or cardiac decompensation can be assumed. Also, acute secondary anemia may lead to penumbral failure and extension of ischemic territory (Bösel et al 2005). Immediate vessel assessment is required to select the best therapeutic strategies.

Despite its proven effectiveness, the use of IV rt-PA thrombolysis remains limited. Initial surveys reported thrombolysis treatment rates for acute stroke patients of 2% in general hospitals and 5% in hospitals with an available stroke unit. Today, most hospitals reach treatment rates of up to 10% and single centers report treatment rates ranking up to 20% (Heuschmann et al 2004, Singer et al 2012) leaving a high proportion of patients without causal treatment options. The major reasons for this were delayed admission, unknown time of symptom onset, minor symptoms, clinical improvement, uncertainty regarding diagnosis, and fear of bleeding complications.

Successful recanalization during systemic thrombolysis depends on the location, length, and composition of the clot (i.e., embolus or thrombus). More distally located clots had the highest recanalization rate in anterior circulation stroke when analyzed with TCD. Complete recanalization occurred in 44.2% of distal MCA occlusions (50 of 113), in 30% of proximal MCA occlusions (49 of 163), and only in 5.9% of terminal ICA occlusions (1 of 17) (Saqqur et al 2007b). Analyzing clot length in 138 patients with acute MCA occlusion undergoing systemic thrombolysis with CT (slice thickness of 2.5 mm) showed successful recanalization in 62 patients in whom the clot length was <8 mm. The median m-RS at hospital discharge was 2. In the remaining 76 patients without recanalization, the clot length was mostly >8 mm. Here patients were discharged with a median m-RS score of 5 (Riedel et al 2011). At least in a rat model, platelet-rich thrombi seemed more resistant to thrombolysis than fibrin-rich clots (Tomkins et al 2015).

The insufficient recanalization rates in proximal and extended occlusions subsequently resulted in intensified clinical use of intra-arterial therapeutic approaches, which had been tested (e.g., in basilar artery (BA) occlusion) even before the establishment of systemic thrombolysis (Zeumer et al 1983). A first approach was the intra-arterial (IA) administration of thrombolytic drug, resulting in higher local concentration and reduced systemic effects, assumed to be more effective and safe. However, the high technical requirements and the procedural delay of treatment onset restricted its use to specialized centers. In the PROACT II (Prolyse in Acute Cerebral Thromboembolism) study, IA urokinase was given within 6 hours of stroke onset. In comparison with heparin alone, substantially better recanalization rates of TICI (Thrombolysis in Cerebral Infarction) 2–3 of 66% versus 18% and a better 90-day clinical outcome were observed (Furlan et al 1999). These results were further confirmed by a meta-analysis of randomized trials published in 2010. There, IA thrombolysis with urokinase or recombinant prourokinase within 6 hours of stroke onset resulted in a better clinical outcome at 3 months compared with systemic thrombolysis. However, at the same time a significant increase of symptomatic intracranial hemorrhages within 24 hours after treatment onset was observed. Despite these results, prourokinase has not yet achieved FDA approval (O’Rourke et al 2010).

All three studies (SYNTHESIS, IMS-III, and MR RESCUE) confirmed at least comparable safety profiles of systemic thrombolysis and endovascular therapy concerning the rates of intracranial hemorrhages, despite not having observed any outcome differences. Long time windows between symptom onset and therapy, the undifferentiated selection of patients, and the use of first- and second-generation devices were considered possible explanations for the unsatisfactory results (Qureshi et al 2014).

Sonothrombolysis

Apart from its use for diagnostic purposes, ultrasound has been shown to have therapeutic potential. Ultrasound is able to cause changes within the fibrin structure of a thrombus by inducing plasma “microstreams.” This subsequently leads to better penetration and action of rt-PA. Both mechanisms accelerate thrombolysis. In 2004, the results of the CLOTBUST (Combined Lysis of Thrombus in Brain Ischemia with Transcranial Ultrasound and Systemic TPA) study were published, which included 126 patients with acute proximal or distal MCA occlusion. The patients were randomized to either rt-PA treatment alone or a combination of IV rt-PA and continuous TCD with a diagnostic 2-MHz transducer over 2 hours. Patients treated with the combined approach had a much higher rate of early recanalization within 2 hours (46% versus 18%) and a trend toward a distinct clinical improvement (29% versus 21%). The trend persisted after 3 months. The study was under-powered, however, to identify a statistically significant difference (Alexandrov et al 2004). A single-center TCCS study in 37 patients with MCA main-stem occlusion also found a higher recanalization rate if continuous ultrasound was applied over 1 hour (46% versus 21%, respectively). However, the number of patients was again too small for a sensitive statistical analysis (Eggers et al 2003).

To improve ultrasound penetration and therefore the thrombolytic potency of ultrasound, the effects of low-frequency ultrasound (300 kHz) were analyzed in acute stroke patients undergoing intravenous thrombolysis. Unfortunately, this approach resulted in a substantially higher number of intracranial hemorrhages and the study had to be terminated early (Daffertshofer et al 2005).

A meta-analysis of six randomized and three nonrandomized clinical sonothrombolysis studies showed that insonation with diagnostic frequencies increased the probability of recanalization threefold and the rate of a better clinical outcome twofold. Rates of symptomatic intracranial hemorrhages did not significantly differ between standard and sonothrombolysis therapies (Tsivgoulis et al 2010). Two other recent meta-analyses concluded that sonothrombolysis reduces 3-month deaths and dependency rates and increases recanalization rates, but does not increase the rate of symptomatic intracranial hemorrhages (Ricci et al 2012, Saqqur et al 2014).

Another potential approach that could enhance ultrasound-induced thrombolysis is the use of air- or gasfilled microbubbles—i.e., ultrasound contrast agents. A small study showed a positive effect using combined TCD ultrasound, IV rt-PA, and microbubbles (Molina et al 2006). A continuative controlled phase IIb TUCSON trial (Transcranial Ultrasound in Clinical SONothrombolysis) reported that a 1.4-mL dose of perflutren-lipid microbubbles did not increase the risk of symptomatic intracranial hemorrhage and could be safely applied during IV rt-PA administration. Using this approach, recanalization rates reached 50–67% (Barreto et al 2009, Molina et al 2009).

Finally, sonothrombolysis has been proposed as a monotherapy, i.e., without additional thrombolytic medications. This is of particular interest specifically as a large proportion of patients are denied IV rt-PA due to contraindications. A pilot study that included patients with MCA main-stem occlusion within a 6-hour time window reported recanalization in 62.5% of patients who underwent 1 hour of continuous TCCS and no recanalization in the noninterventional group (Eggers et al 2005).

One limitation for the widespread use of sonothrombolysis may be the lack of a sufficient temporal bone window for MCA insonation as the main target vessel. In a study of 179 patients (42% female), 39% had an M1- or M2-MCA occlusion. Overall, 23% had inadequate insonation conditions. Remarkably, a comparable number (18%) were ineligible for CTA or MRA diagnostic evaluation because of contraindications or severe motion artifacts (Nolte et al 2013). A second important limitation is the operator dependency in ultrasound examinations. As the availability of experienced sonographers in the emergency department is currently limited to a few centers (Tsivgoulis and Alexandrov 2008), the development of an investigator-independent “hands-free” 2-MHz trans cranial ultrasound device seems an attractive option. Following promising results from smaller phase IIa/IIb studies, a hands-free device was studied in the randomized controlled multi-center CLOTBUT-ER trial (Barlinn et al 2013, Barreto et al 2013, Schellinger et al 2015).

Angiologic and Anatomic Aspects

Not all patients with clinical stroke symptoms have a major brain-supplying artery occlusion, or even an occlusion that could potentially be reopened by thrombolysis. This patient group, however, is exposed to the 5–10% risk of an intracranial bleed. A major argument against an extended vascular diagnosis has been the time factor. But with the increasing availability of multimodal CT, MRI, and ultrasound techniques, which permit a fast and reliable diagnosis of vessel occlusions, the vascular status can now be obtained without significant delay. This in turn will help to stratify patients early, and subsequently allow for the administration of individualized treatment strategies, likely leading to better clinical outcomes.

Perfusion and diffusion-weighted MRI allows brain tissue that is functionally but not yet structurally impaired (penumbra) to be visualized, defining the penumbra as the difference between the perfusion MRI and diffusion MRI. From this technique the so-called “mismatch concept” was developed and has been successfully used to select patients for thrombolysis and to prolong the time window of IV rt-PA treatment to up to 6 hours and longer in selected cases (Hacke et al 2009).

A similar approach is followed in current CT perfusion studies. The CT technique has a particular logistic advantage as CT access for emergency patients is usually easy to obtain. Using CT, the ischemic penumbra or “tissue at risk” can be defined as the difference between the CBF (analogous to perfusion MRI) and the CBV (analogous to diffusion MRI). A comparative study in 42 stroke patients showed that tissue at risk, as determined by perfusion CT and CTA, was equivalent to the MRI results in all but one case (Wintermark et al 2007). Moreover, recent studies support the suitability of perfusion CT for stroke detection (Campbell et al 2015, Hana et al 2014) including a meta-analysis that confirmed perfusion CT as having a high sensitivity and a very high specificity for detecting infarcts (Biesbroek et al 2013).

Finally, ultrasound perfusion tests have also been performed in acute stroke patients. Ultrasound follows the same basic approach as MRI and CT, i.e., it analyzes the perfusion kinetics of a contrast bolus within the micro-circulation. However, for several reasons, this technique is currently still experimental and rather limited (Eyding et al 2006, Kern et al 2011, Meairs and Kern 2015, Seidel et al 2013).

Digital subtraction angiography (DSA) remains the reference method for the evaluation of intracranial occlusions. In clinical practice, multislice CTA, yielding similar results, is being increasingly used. In BA pathology, CTA sensitivity is even higher as it has been shown to detect distal BA near-occlusions that were considered complete occlusion using DSA. After correction for false-positive BA occlusions, sensitivity, specificity, and positive and negative predictive values for the detection of intracranial occlusions of the major arteries by CTA were all 100%. Respective values for time-of-flight (TOF)-MRA in the same patient group were 87% sensitivity, 98% specificity, 59% positive predictive value, and 99.5% negative predictive value (Bash et al 2005). When analyzing more distal artery occlusions the limitation of TOF-MRA becomes more evident. Using contrast-enhanced MRA 20% of main-stem arteries or their branches were patent arteries, which had been interpreted as occluded using TOF-MRA (Yang et al 2002). The spatial resolution of TOF-MRA can be improved if 3-T and sensitivity-encoding techniques are combined, which has been shown to improve diagnostic results. Compared with DSA, sensitivity, specificity, and positive and negative predictive values of 100%, 99%, 87%, and 100% were reported but the authors included only MCA and ICA occlusions (Choi et al 2007).

CT and MRI can mainly be used as “single shot” methods. They are too laborious, too expensive, or too stressful for the patient to be used for continuous monitoring. In comparison, ultrasound has the advantage of permitting serial as well as continuous measurements without these restrictions. In the hands of an experienced sonographer, TCD may allow diagnosis of main-stem occlusion, at least for the main stem of the MCA. In acute stroke patients, however, TCCS is the superior method provided that the transcranial bone window is sufficient or signals are enhanced by contrast agents. A proximal MCA occlusion can be diagnosed simply if the color signal and Doppler spectrum in projection of the artery in question are absent. In a small study (10 stroke patients with a main-stem MCA occlusion diagnosed using TOF-MRA), TCCS confirmed the diagnosis of a main-stem MCA occlusion in all patients (Kenton et al 1997). In a second study 20 of the 23 patients who had vascular pathology and underwent contrast-enhanced TCCS were correctly diagnosed, whereas TCD and unenhanced TCCS were used successfully only in 14 and 7 patients (Goertler et al 1998). In hyperacute stroke, the diagnostic accuracy of TCCS in comparison to CTA and MRA was analyzed in 58 stroke patients within the first 6 hours of stroke onset by unenhanced and, if necessary, subsequent contrast-enhanced extra- and intracranial duplex ultrasound. Examinations were performed before the initiation of other diagnostics while patients were still in the emergency department. Echo contrast agent use was considered necessary in 51 patients. The mean duration of the complete examination was 21.3 minutes in the combined native and contrast-enhanced examination and 13.6 minutes if native examination was performed alone. Ultrasound findings were confirmed using DSA in 31 of 32 patients (Gerriets et al 2002). Shorter examination times could probably be achieved if contrast agents were administered at the beginning of the ultra-sound examination. In an own clinical ultrasound study we analyzed the time needed for the duplexsonographic assessment of all extra- and intracranial arteries (extra-cranial ICA and VA, intracranial MCA, ACA, PCA, V4-VA, and BA) if reduced items were assessed (vessel status: open, occluded, stenosis >50%, distal stenosis/occlusion) and if intravenous echo contrast (SonoVue) was initially used. Analysis of 126 patients revealed a mean insonation time of ~6 minutes with a range of 4–12 minutes (Schreiber et al 2010).