Recanalization after thrombolysis in stroke patients

IV thrombolysis with tissue plasminogen activator (tPA) is the first-choice therapy for ischemic stroke if applicable within 3 hours.^1,2 There is broad consensus that thrombolysis improves clinical outcome by resolving the occlusive clot and restoring adequate blood flow.² The rate of successful recanalization after IV tPA, however, is not well established because the large intervention trials have not systematically assessed vessel status before and after therapy. The few data available are from a limited set of stroke databases and case series,^3–20 most of which were comparatively small and from secondary referral centers. Moreover, these studies were partly retrospective in design, and some applied continuous TCD monitoring,^{3–6,12,13,15} which, by itself, may affect recanalization.

The current prospective study aims at assessing the frequency of recanalization after IV thrombolysis in a series of 64 consecutive and unselected patients with documented occlusion of the intracranial internal carotid artery (ICA) or middle cerebral artery (MCA). An additional focus will be on the identification of predictors of vessel recanalization and on its prognostic potential.

Methods.

Results.

We evaluated 64 consecutive patients (36 female, 28 male) with anterior circulation stroke treated with IV tPA. Demographic factors, risk profiles, and baseline clinical findings are summarized in table 1. Mean age was 65.7 years, and median baseline NIHSS was 18 compared with 14 (National Institute of Neurological Disorders and Stroke), 13 (European Cooperative Acute Stroke Study [ECASS] I), 11 (ECASS II), and 9 (Alteplase Thrombolysis for Acute Noninterventional Therapy in Ischemic Stroke [ATLANTIS]).¹

Twenty-four hours after administration of IV tPA, complete recanalization was achieved in 36 patients (56.3%), whereas partial or no recanalization was observed in 10 (15.6%) and 18 patients (28.1%). Classifications based on CTA and TCD were in close agreement (κ coefficient = 0.84), and recanalization rates were identical irrespective of ascertainment by CTA or TCD.

The frequency of recanalization in various subgroups is summarized in figure E-1 on the Neurology Web site (www.neurology.org). Significant differences were found between subjects with and without diabetes (p < 0.001) and with and without coexisting extracranial carotid occlusions (p = 0.03) (figure 1). Moreover, the probability of vessel reopening decreased with advancing age, higher baseline NIHSS score, and a more proximal site of vessel occlusion. Neither of these trends, however, reached significance. Reopening was obtained in 62.5% (10 of 16) of the occluded extracranial carotid arteries.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Frequency of recanalization in the overall study population and in subgroups with significant differences in recanalization rates. Values are percentages. ICA = internal carotid artery.

Complications of IV tPA were rare, and symptomatic intracerebral bleeding occurred in a single case with persistent vessel occlusion 24 hours after therapy. Six patients developed large MCA infarcts and died within 10 days after thrombolysis.

At day 90, median mRS was 3.0 (range, 0 to 6). A total of 23 of the 64 patients (36%) became functionally independent (mRS ≤ 2). Clinical outcome was substantially better in patients with complete vessel recanalization than in those with partial or no recanalization (figure 2). Further univariate predictors of a favorable clinical outcome were younger age (p = 0.007), low NIHSS (p = 0.003), and absence of diabetes (p = 0.005). A multivariate logistic regression analysis allowing for all variables identified five independent predictors (figure 3). Vessel recanalization, age, and NIHSS maintained significance, diabetes fell short of significance because of its strong association with recanalization, and aspirin pretreatment and onset-to-needle time newly entered the model. Predictive significance of the latter variable was masked in the univariate analysis by its strong association with NIHSS score and age.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. Probability of a favorable clinical outcome (modified Rankin score [mRS] ≤ 2) in patients with complete, partial, and no recanalization. Values are percentages.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3. Predictors for a favorable clinical outcome (modified Rankin score [mRS] ≤ 2) derived from a multivariate logistic regression analysis. Odds ratios and 95% CIs were calculated for a 5-point increase in the NIH Stroke Scale (NIHSS), a 10-year increase in age, and a 30-minute increase in the onset-to-needle time. Patients with partial and no recanalization were combined and compared against patients with complete recanalization.

Discussion.

The rate of recanalization achievable by standard IV thrombolysis of anterior circulation stroke is not well established, mainly because the large trials that qualified thrombolysis as a first-choice treatment for stroke did not systematically assess vessel status before and after intervention. Recently, a number of studies focusing on the concept of continuous TCD-enhanced thrombolysis^{3–6,12,13,15} reported recanalization rates of 30% to 56% (complete recanalization) and 60% to 71% (any recanalization) occurring within 2 to 6 hours after thrombolysis (table 2). These data, however, are not easily transferred to standard stroke patients because of trial-specific inclusion and exclusion criteria and putative effects of continuous ultrasound on recanalization. Only a few comparatively small studies have assessed recanalization rates after thrombolysis in the clinical routine setting.^{7–11,16–19} Table 2 gives a summary of these evaluations, part of which were retrospective in design and conducted in secondary referral centers. Rates ranged from 17% to 46% for complete recanalization and 31% to 92% for any recanalization. Our prospective study included 64 consecutive thrombolysis patients with documented occlusion of MCA, ICA, or both, yielding a 24-hour rate (95% CI) of complete recanalization of 56.3% (44.1% to 68.5%). In a previous prospective multicenter study, partial or complete recanalization after IV tPA was observed in 34.4% of the 93 patients enrolled. The low rate compared with our study may be explained by the design (dose escalation), mode of drug administration (no bolus), an onset-to-needle time clearly beyond the 3-hour window (mean, 5.4 hours), and assessment of vessel reopening 1 hour after therapy.²⁰

In our study, the probability of recanalization decreased with a more proximal site of vessel occlusion. This tendency, however, was not significant, and rates for carotid T occlusion still amounted at 46.2%. In keeping with our findings, two recent studies suggested that patients with carotid T occlusion and IV thrombolysis have a better outcome and more frequent vessel reopening than assumed previously.^8,19 Little information is currently available on how often occlusions of the extracranial carotid artery recanalize after IV thrombolysis.²⁰ The frequency in our series was high, at 62.5% (10 of 16 patients). However, in most instances high-grade stenosis persisted, and no effect on clinical outcome was found.

Our study identified two independent predictors of recanalization. First, the rate of intracranial vessel reopening was lower in patients with a coexisting extracranial carotid occlusion. Secondly, and most importantly, IV thrombolysis did not work well among stroke patients with diabetes. Actually, rates of complete recanalization were as low as 9.1% as opposed to 66.0% in nondiabetics (p < 0.001). This finding may partly explain the worse clinical outcome in this group of patients^12,22 and is in close agreement with a previous investigation suggesting a low efficiency of thrombolysis among diabetics with ST elevation myocardial infarction.²³ There are various potential clues of explanation for this phenomenon, but it is tempting to speculate about a central role of plasminogen activator inhibitor (PAI). Levels of PAI are substantially elevated in diabetic patients, and this abnormality is not reversed by standard therapy.²⁴ Actually, baseline levels of PAI have been shown to be inversely related to clinical outcome in thrombolysis patients,²⁵ and high PAI may cause thrombolysis failure.²⁶ Inhibitory effects of PAI on tPA action have consistently been shown in experimental and in vivo (animal) studies.^17,25,26 It should also be mentioned that diabetes impairs the benefit of primary coronary angioplasty on long-term clinical outcome in ST-segment elevation myocardial infarction patients with diabetes,^27,28 and diabetes and even baseline hyperglycemia in the absence of diabetes generally are unfavorable predictors of clinical outcome in stroke patients.

The low recanalization rate among diabetics in our study must be interpreted in light of the limited number of patients with diabetes and, thus, awaits confirmation in larger studies. Provided that this alarming observation holds true in future research, improvement of thrombolysis efficacy in the large segment of diabetics may be viewed as an urgent challenge in acute stroke therapy. Recent advances in thrombolysis potentially relevant to this particular issue are the development of new thrombolytics and an enhancement of standard therapy by novel techniques such as sonothrombolysis, combined intra-arterial thrombolysis, and mechanical clot fragmentation.

Adequate recanalization was among the strongest predictors of a favorable clinical outcome at day 90 after stroke (58.3% vs 10.0% and 5.6% in subjects with partial or no recanalization, p < 0.001). This finding fits very well to the penumbra concept and proves that the advantage of reperfusion is not offset by a higher risk of bleeding. Actually, none of the patients with documented vessel recanalization developed a symptomatic intracerebral hematoma. Of note, successful recanalization 24 hours after therapy predicts a favorable outcome on top of the information provided by the variables of age, NIHSS, and onset-to-needle time.^3,6,11,29

Our study has several strengths, including the prospective design, standardized collection of data, and the fact that the whole survey area was served by a single stroke unit. The latter issue is very important because it provides a safeguard against the usual selection in secondary referral centers. There are potential limitations as well. First, both CTA and TCD were used to assess vessel occlusion and recanalization. In cases in whom both modalities were performed, however, there was adequate accordance, as has been reported previously.²¹ Second, we could not obtain 24-hour follow-up vessel images in 10 patients. Six of these patients died. In the other four, 24-hour TCD or CTA were obtained more than 24 hours after thrombolysis and documented successful vessel reopening in all of them. Third, the study design did not allow definition of the exact time to recanalization, which may be relevant to individual prognosis.^3,6,11,29 The fact that onset-to-treatment time had no apparent effect on the overall rate of recanalization but emerged as a predictor of clinical outcome strongly suggests that the time of recanalization and subsequent salvage of the penumbra made the difference. Finally, the number of patients studied is comparatively small for complex analyses. Accordingly, in the multivariate logistic regression analysis focusing on risk predictors for a favorable clinical outcome, confidence intervals of risk estimates were broad, and the power to detect associations was limited. A low number of subjects relative to the number of variables tested may also give rise to chance findings.