Thrombolysis for Ischemic Stroke Associated With Infective Endocarditis
Results From the Nationwide Inpatient Sample
Background and Purpose—Cerebral ischemic events are highly prevalent and associated with high rates of death and disability in patients with infective endocarditis (IE). However, the role of thrombolysis in these patients remains unclear. We sought to determine the rates and outcomes of acute ischemic stroke patients with IE treated with intravenous thrombolysis (IVT).
Methods—We determined the rates of post-thrombolytic intracerebral hemorrhage and favorable outcome among acute ischemic stroke patients with IE treated with IVT. Patients were identified using Nationwide Inpatient Sample data from 2002 to 2010. We compared the rates of various outcomes with ischemic stroke patients without IE treated with IVT.
Results—There were 222 patients (mean age 59±18 years; 46% women) who were treated with IVT for acute ischemic stroke associated with IE and 134 048 patients (mean age 69±15 years; 49% women) who were treated for stroke without IE. The rate of post-thrombolytic intracerebral hemorrhage was significantly higher in patients with IE compared with those without IE (20% versus 6.5%; P=0.006). There was a significantly lower rate of favorable outcome in the IE group (10% versus 37%; P=0.01).
Conclusions—High rates of post-thrombolytic intracerebral hemorrhage and low rates of favorable outcome mandate caution in using IVT in acute ischemic stroke patients with IE.
Cerebral ischemic events are highly prevalent in patients with infective endocarditis (IE) and associated with high rates of death and disability, ranging up to 73%.1–5 Patients with acute ischemic stroke (AIS) are treated with intravenous thrombolysis (IVT) within 4.5 hours of symptom onset as recommended by American Heart Association/American Stroke Association guidelines.6 However, previous trials, including the National Institutes of Neurological Disorders and Stroke thrombolysis study and the Prolyse in Acute Cerebral Thromboembolism II trial, have excluded patients presenting with AIS associated with IE.1,4,7 The issue is further complicated by the fact that the diagnosis of IE may not be evident at the time of the acute ischemic event. Case reports of patients treated with IVT in the setting of AIS caused by IE have shown mixed results.2,7,8 We sought to determine the rates and outcomes of AIS patients with IE treated with IVT using nationally representative data.
We used data files from the Nationwide Inpatient Sample from 2002 to 2010. We used the International Classification of Disease, Ninth Revision, Clinical Modification for ischemic stroke, IE, and the use of thrombolysis to identify patients. Outcomes were defined as rates of post-thrombolytic intracerebral hemorrhage (ICH) and favorable outcome (discharge disposition of home/self-care). Univariate and multivariate analyses were performed to determine the effect of IE on rates of post-thrombolytic ICH and favorable outcome among survivors. For the complete methodology of our study, refer to the Methods in the online-only Data Supplement.
There were 222 patients (mean age 59±18 years; 46% women) who were treated with IVT for AIS secondary to IE and 134 048 patients (mean age 69±15 years; 49% women) who were treated for AIS without IE (Table). The rate of post-thrombolytic ICH was significantly higher in patients with IE compared with non-IE AIS patients (20% versus 6.5%; P=0.006). There was a significantly lower rate of favorable outcome in the IE group (10% versus 37%; P=0.01) when compared with the non-IE AIS treated group. The length of hospital stay was significantly longer among thrombolytic-treated patients with IE (P=0.006) and the mean hospital charges were significantly higher (P=0.01). Refer to our Results in the online-only Data Supplement for further analysis.
Anderson et al9 noted in their study that 35% of stroke patients with IE died during their acute admission, and a total of 52% died within 1 year. With this high rate of mortality among patients with AIS associated with IE, it is important to determine whether current performance of thrombolysis for AIS is beneficial within this unique group of patients. Current American Heart Association/American Stroke Association guidelines do not exclude eligible AIS patients with IE from receiving IVT.
The use of thrombolytics is controversial because of high rates of spontaneous hemorrhagic transformations of infarctions among patients with IE.3,4,8 The high rate is partly attributed to concurrent existence of mycotic aneurysms with subsequent rupture. Other potential pathogeneses include the presence of pyogenic arteritis, microabscesses, immune complex–mediated arteritis, and infiltration of meningeal vasculature.2
We found that the rates of post-thrombolytic ICH were significantly higher and the rates of favorable outcomes were significantly lower within the IE group when compared with the non-IE group. Recent case reports have shown variable rates of favorable and unfavorable outcomes associated with IVT use in patients with AIS associated with IE.2,7,8 Our study had a relatively large number of patients with IE and, therefore, was devoid of biases introduced by patient referral patterns and socioeconomic factors within catchment areas seen in single center studies. The precision of estimates may be higher in our study for post-thrombolytic ICH and favorable outcomes. In the absence of a randomized comparison between thrombolytic and nonthrombolytic treatment among AIS patients with IE, thrombolytic treatment cannot be prohibited in this group of patients. The results do support a cautious approach with careful risk–benefit analysis on a case-by-case basis. The results also emphasize the need for early diagnosis of IE in AIS patients given the prognostic significance among thrombolytic-treated patients. In patients with concurrent fever and leukocytosis, documentation of endocardial involvement by auscultation and emergent echocardiography may identify such patients in the acute setting when results of blood cultures may not be available.
We recognize that our study has certain limitations. Inaccuracies with coding can affect patient identification and result in underestimation of proportion of patients in the groups created for comparison. International Classification of Disease, Ninth Revision, Clinical Modification codes for ischemic stroke (when listed as primary diagnosis) and thrombolysis have high specificity but moderate sensitivity and, thus, underestimate the prevalence.10 The procedure code 99.10 does not differentiate between various thrombolytic agents and approaches, and it is possible that some patients may not be treated with IV alteplase. Furthermore, we cannot comment on the appropriateness of use and protocol violations. Although the accuracy of International Classification of Disease, Ninth Revision, Clinical Modification code for IE is unknown, a relatively high-positive predictive value is expected because the codes are necessary for reimbursement of posthospitalization IV antibiotic treatment. However, the criteria used for diagnosis may be varied between institutions. We are also unable to determine the severity of neurological deficits, baseline functioning of patients, and underlying location (lacunar versus nonlacunar) and cannot exclude the possibility that IE patients had more severe deficits and preexisting disability. We were able to ascertain presence of ICH using International Classification of Disease, Ninth Revision, Clinical Modification codes, but were unable to categorize ICH based on severity. Although a formal outcome scale was not used, the use of discharge home as an outcome has a negative predictive value of 95% for modified Rankin scale scores ≥2 at 3 months.11 Our data were derived from several institutions, including rural, urban teaching, and urban nonteaching hospitals. Institutional characteristics, such as availability of dedicated neuro-critical care and stroke services, can lead to differences in patient outcomes and our results may be underrepresenting the care and outcomes at comprehensive stroke centers. The difference in length of stay must be interpreted with caution because of several posthospitalization factors, such as need for postdischarge IV antibiotic treatment and bed availability at destination.
The high rate of ICH and low rates of favorable outcome mandate caution in using IVT in AIS patients associated with IE. Endovascular treatment may be evaluated in future studies because of lower thrombolytic doses and concurrent angiographic assessment of medium-sized intracranial arteries.
Dr Qureshi has received funding from the National Institutes of Health U01-NS062091-01A2 (medication provided by EKR Therapeutics), American Heart Association Established Investigator Award 0840053 N, and Minnesota Medical Foundation, Minneapolis, Minnesota. The other authors have no conflicts to report.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.113.001602/-/DC1.
- Received March 25, 2013.
- Revision received June 17, 2013.
- Accepted July 3, 2013.
- © 2013 American Heart Association, Inc.
- Del Zoppo GJ,
- Saver JL,
- Jauch EC,
- Adams HP Jr.
- Sontineni SP,
- Mooss AN,
- Andukuri VG,
- Schima SM,
- Esterbrooks D