HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Irving L. Kron Curtis G. Tribble Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Parrino, P. E. Articles by Tribble, C. G. Search for Related Content PubMed PubMed Citation Articles by Parrino, P. E. Articles by Tribble, C. G.

Ann Thorac Surg 1999;67:59-64
© 1999 The Society of Thoracic Surgeons

---

Original Articles

Does a focal neurologic deficit contraindicate operation in a patient with endocarditis?

^a Division of Thoracic and Cardiovascular Surgery, Department of Surgery, University of Virginia Health Sciences Center, Charlottesville, Virginia, USA

Address reprint requests to Dr Tribble, Department of Surgery, University of Virginia Health Sciences Center, Box 181-95, Charlottesville, VA 22908
e-mail: cgt2e{at}virginia.edu

Presented at the Poster Session of the Thirty-fourth Annual Meeting of The Society of Thoracic Surgeons, New Orleans, LA, Jan 26–28, 1998.

	Abstract

Top Footnotes Abstract Introduction Material and methods Results Comment References

 
Background. As many as 40% of patients with left-sided bacterial endocarditis will sustain a neurologic insult. The importance of a neurologic change as an indication or a contraindication for valve replacement remains controversial.

Methods. We performed a retrospective analysis of the records of 33 patients admitted to the University of Virginia Health Sciences Center between January 1, 1978, and June 30, 1996, with a diagnosis of endocarditis and a neurologic change.

Results. All 33 patients had echocardiographic or pathologic evidence of left-sided endocarditis; 23 were seen with focal neurologic findings and had a mortality rate of 22% (5 of 23), and 10 patients were seen with nonfocal, diffuse encephalopathy and had a mortality rate of 60% (6 of 10) (p < 0.05). Of the 33 patients, 14 underwent operation and 19 were treated medically. The mortality rate was 21.4% (3 of 14) in the surgical group and 42.1% (8 of 19) in the medical group (p = not significant). In 71% (10 of 14) of the surgical patients, the operation was done within 1 week of the neurologic event. Additional neurologic deterioration occurred in 18.2% (2 of 11) of survivors in the surgical group and 9.1% (1 of 11) in the medical group (p = not significant).

Conclusions. Choosing therapy for a patient with endocarditis and a neurologic change remains a difficult challenge. Initial findings of nonfocal, global dysfunction on examination are a predictor of a poor outcome. By comparing surgical and medical groups derived from the same series of patients, it is clear that patients with bacterial endocarditis and central nervous system changes face substantial mortality regardless of intervention. However, these data demonstrate that when compared with a similar group of medical patients, surgical patients who require and receive operation early in the course of their illness do comparatively well. Improving outcomes by delaying surgical intervention may serve to "select out" hardier patients but will lead to the death of patients who might benefit from such intervention.

	Introduction

Top Footnotes Abstract Introduction Material and methods Results Comment References

 
Bacterial endocarditis (BE) continues to be a major clinical problem. Despite advances in medical and surgical therapy, mortality remains 30% in some subsets of affected patients [1]. Death can occur secondary to systemic effects of the infection, such as emboli, the persistence of resistant organisms, a persistent septic state, or as a result of hemodynamic collapse brought on by direct injury to intracardiac structures. Neurologic complications can occur in up to 40% of patients, and the presence of such a complication can increase the mortality rate to as high as 58% [2]. The management, including the timing of surgical intervention, of patients with BE and central nervous system (CNS) complications is difficult. Replacement of failing valves early in the course of complicated BE is associated with improved survival; however, this must be weighed against the risks of cardiopulmonary bypass in patients with known intracranial lesions [3, 4].

Several recent studies [5–7] examining the timing of operation in patients with BE and CNS complications have developed recommendations based on outcome comparisons among surgical patients. These data, however, do not allow for comparison with patients with BE and CNS complications managed nonsurgically. The purpose of this retrospective study was to analyze outcomes for both medically and surgically managed patients with BE and CNS complications seen at our institution. By comparing these groups drawn from the same demographic and referral area, we attempted to discern differences in outcome based on both the presence and the timing of surgical intervention.

	Material and methods

Top Footnotes Abstract Introduction Material and methods Results Comment References

 
We reviewed the records of all patients admitted to the University of Virginia Health Sciences Center between January 1, 1978, and June 30, 1996, with a diagnosis of BE and an acute neurologic change. Only patients with left-sided BE were included. Criteria for the diagnosis of neurologic change were based solely on the results of the physical examination; patients with lesions on computed tomographic scan, patients with transient ischemic attacks, and patients with headaches whose examination results were normal were excluded. Neurologic change was defined as the presence of a focal neurologic deficit or an acute change in mental status. These changes were described as either focal or nonfocal findings. We identified 33 patients with these criteria for inclusion in this study.

Data collected on the 33 patients included age, sex, presenting signs and symptoms, type of neurologic deficit, time between admission and CNS change, cranial imaging results, echocardiographic procedures and results, blood culture results, surgical procedures and timing related to both admission and onset of CNS change, indications for operation, cause of death, additional complications including secondary neurologic events, and neurologic status at discharge.

All data are expressed as the mean ± the standard error of the mean except where otherwise noted.

	Results

Top Footnotes Abstract Introduction Material and methods Results Comment References

 
In the study group of 33 patients, there were 25 men (76%) and 8 women (24%) with a mean age of 48.8 ± 3.4 years. Twenty-two patients (67%) were seen with acute CNS changes. Four patients had development of CNS changes while being treated for endocarditis, 5 with BE and acute CNS changes were transferred from other institutions, 1 patient ignored his symptoms for 13 days until he was seen, and 1 patient had development of nosocomial endocarditis while being treated for Evans’s syndrome (acquired hemolytic anemia and thrombocytopenia). At admission, 21 patients (64%) had a fever, 8 (24%) had a new murmur, 19 (58%) had malaise, and 14 (42%) had evidence of peripheral emboli. Only 3 patients (9%) had a history of intravenous drug abuse. All patients met the criteria of Von Reyn and associates [8] for definite or probable endocarditis.

Twenty-three patients (70%) demonstrated focal neurologic deficits, and 10 (30%) had nonfocal, diffuse encephalopathy. All (97%) but 1 patient underwent cranial imaging. Of these 32 patients, 23 (72%) demonstrated lesions. Twenty (87%) of these lesions were embolic, and three (13%) were hemorrhagic. Four patients (12%) were found to have mycotic aneurysms, two of which ruptured and were subsequently clipped. In the patients with nonfocal findings, 6 (60%) had lesions on imaging, and 3 (50%) of these patients had multiple infarcts. In the group of patients with focal findings, 17 (74%) had lesions, and only 2 (12%) demonstrated multiple infarcts.

All patients had echocardiography, with 22 (67%) demonstrating lesions on the initial examination. Of the 11 with negative initial results, 6 (55%) were later found to have lesions, either by transesophageal echocardiography or by repeat transthoracic examination, resulting in an overall 85% rate of echocardiographic evidence of lesions. Table 1 lists the various valves and combinations of valves found to be involved as well as the estimated size of vegetations. Five patients (15%) had prosthetic valve endocarditis; four were porcine valves, and one was a mechanical valve. Only one case of endocarditis occurred within 60 days after the valve replacement procedure.

	Comment

Top Footnotes Abstract Introduction Material and methods Results Comment References

 
Despite advances in both medical and surgical care, the mortality associated with BE remains high. In a study published in 1989 [9] that used a relatively broad approach to diagnosing BE, the mortality rate remained near 14%. Several authors [3, 10, 11] have demonstrated improved results with early surgical intervention for the treatment of complicated BE. These algorithms become more difficult to apply when the clinical picture is clouded by the presence of neurologic changes. This study has focused on timing of surgical intervention and comparing outcomes for surgically treated patients versus those managed nonsurgically.

Neurologic changes can occur in as many as 40% of patients with BE and range from stroke or transient ischemic attack, toxic encephalopathy, meningitis, brain abscess, visual loss, seizures, or headache to acute mononeuropathy [2, 12]. In one series [13], 29% of patients with BE were seen with a chief complaint of a focal neurologic deficit, altered mental state, or headache. The presence of more severe events, such as stroke, can raise the mortality rate associated with BE to 58% [2].

"Stroke" as a physical finding can be caused by a number of events. However, the most worrisome when considering a cardiac surgical intervention is a hemorrhagic infarction. Factors contributing to the development of hemorrhagic infarctions in the setting of BE include pyogenic arteritis, hemorrhagic transformation of an embolic infarction, or rupture of a mycotic aneurysm [14]. The presence of a hemorrhagic infarction substantially increases the risk of postoperative stroke and death [6].

In our study, there were 3 patients with hemorrhagic infarctions. In 1 patient, the diagnosis was hemorrhagic transformation of an embolic infarction (the cerebral arteriogram did not demonstrate an aneurysm), and brain death occurred 4 days later. The second patient had a mycotic aneurysm clipped 7 days after admission and several months later, had repair of a congenital ventricular septal defect. This patient was not included in the surgical group because of the long interval between the CNS event and operation. The third patient had an aneurysm clipped 8 days after admission and died 26 days later of multisystem organ failure.

The group of patients in this study is comparable to groups in other series. The mean age of our patients was 48.8 years, and the male to female ratio was 3:1. This is similar to what is reported in other series [6, 15, 16]. The prevalence of Staphylococcus aureus was 43%, which is comparable to the range of 38% to 52% in other studies [7, 17]. The involved valves in this study were equally divided between the mitral and aortic, which is slightly different from other series, in which the aortic valve has predominated [3, 6, 7, 16]. The overall mortality rate in this series was 33%, with a 21% mortality rate in the surgical group and a 42% rate in the medical group. This finding also is comparable to the mortality rates reported in other series [15, 16].

The management of patients with BE and a CNS change is complex. We followed a fairly aggressive course in our management of these patients, as is reflected in Tables 3 and 4. More than half (57%) of the patients in the surgical group had operation to replace infected and failing valves within 72 hours of the onset of the CNS change. No patient in the medical group had any of the standard indications for urgent operation. Only 2 patients in the medical group had congestive heart failure, and in each, it was described as mild. Both of these patients also required clipping of a mycotic aneurysm after intracerebral hemorrhage. One patient died soon thereafter of multisystem organ failure, whereas the other ultimately had repair of a congenital ventricular septal defect. Therefore, few patients with surgically correctable lesions were excluded from surgical consideration because of other factors. The indications for operation in this group of patients were the following: heart failure, 64.3% (9 of 14); recurrent systemic emboli, 21.4% (3 of 14); prosthetic valve endocarditis, 7.1% (1 of 14); and atypical organism, 7.1% (1 of 14).

The difference in mortality between the surgically and medically treated groups was not significant (21% versus 42%; p = 0.213), despite the fact that surgical intervention was carried out within 7 days in 10 (71%) of the 14 patients, which has been shown to correlate with increased surgical risk [5, 7]. Although all of the deaths in the surgical group occurred in patients treated operatively within 24 hours of the neurologic event, the interval between this event and death was not significantly different from the same interval in the medically treated patients (see Table 6). Therefore, this mortality likely represents the natural course of disease in this group of critically ill patients. Some studies have recommended a delay, when possible, in operating on patients with BE and a CNS event, though they included only patients who underwent operation [5–7].

By including a medical group of patients drawn from the same population, this study demonstrates that a high mortality can be expected in both medically and surgically treated patients. There are other important differences between this study and previous reports. Gillinov and associates [7] excluded patients with encephalopathy, which has been shown to be the second most common CNS finding in patients with BE [12, 16]. These patients were included in our study, and in fact, encephalopathy was found to have a significant predictive value for mortality; 60% of such patients died versus 22% of the patients with focal changes on examination (p < 0.05). The increased mortality in the group with nonfocal, diffuse encephalopathy reflects the presence of patients with more severe illness probably due to a higher number of microemboli, resulting in global dysfunction without localizing signs [12]. Two of the 3 surgical patients in this series who died had nonfocal, diffuse encephalopathy at admission. Ting and coworkers [6], in a review of 45 patients, found only preoperative hemorrhage to be associated with mortality and postoperative stroke. However, a large percentage of the patients in that study were asymptomatic and were included on the basis of the preoperative computed tomographic scan rather than physical findings. Eishi and colleagues [5] evaluated the timing of operation after CNS events in 181 patients with BE. No patients with encephalopathy were included; in addition, it is unclear whether or not patients were asymptomatic or symptomatic. All patients in our study were symptomatic. In summary, Eishi and coauthors [5] and Gillinov and associates [7] found that a delay in operation for BE resulted in a lower complication rate. Our data, however, support the conclusions of Ting and colleagues [6]. The timing of surgical intervention after stroke in patients with ischemic events does not appear to significantly affect outcome.

In this study, we compared two groups derived from the same series of patients. One group received surgical intervention based on the standard criteria for surgical therapy in BE, whereas the other group did not require surgical therapy and was treated medically. No patient in the medical group was denied operation on the basis of neurologic status, as reflected in the low number of patients with congestive heart failure in that group. The difference in mortality between the groups was not significant, nor were the differences in in-hospital complications (see Table 4). When the entire group of patients was divided with respect to type of neurologic change (focal versus nonfocal findings), a significant difference in mortality was seen (22% versus 60%, respectively; p < 0.05).

Patients with BE and CNS changes on examination represent a critically ill group of patients with significant risk of death. Nonfocal, diffuse encephalopathy as a presenting neurologic sign is a predictor of even higher mortality than that anticipated in these patients. The cardiac surgical evaluation of patients with BE and CNS changes should include a cranial computed tomographic scan, as this has clearly been shown to be helpful in the diagnosis of intracranial lesions, specifically hemorrhage.

Patients with focal deficits and computed tomographic scans that do not demonstrate hemorrhagic lesions should be strongly considered for operation when referred for evaluation. If operation is indicated because of the presence of heart failure, ongoing sepsis, recurrent emboli, infection resistant to antibiotic therapy, or other findings that mandate surgical intervention, expeditious operation should be offered. These patients, even when surgically treated early in the course of the illness, do at least as well as their medically treated counterparts and may even fare better.

Patients with nonfocal findings and computed tomographic scans that do not demonstrate hemorrhage are at increased risk of additional neurologic events and death compared with patients with focal changes. However, in this series, the patients with nonfocal changes who were treated surgically had nearly the same mortality rate as those treated medically (66% versus 57%, respectively; p = not significant). Thus, although the prognosis is grave regardless of therapy, these patients should also be considered for surgical intervention when referred for evaluation.

Patients with intracranial hemorrhage represent a subset of patients who are at high risk of death regardless of the therapy initiated. Though there were only 3 patients in this series with hemorrhagic infarctions, other reports [6] have demonstrated a high rate of complications and death associated with acute cardiac operation in such patients. The presence of a hemorrhagic infarction probably precludes cardiac surgical intervention in all but extreme cases, as acute surgical therapy may increase mortality in this group beyond what would be expected in patients undergoing medical therapy only.

In this study of surgical and medical groups derived from the same series of patients, it is clear that patients with BE and CNS changes face significant risk of mortality regardless of intervention. Nevertheless, these data demonstrate that when compared with a similar group of medical patients, surgical patients who require and receive operation early in the course of their illness do comparatively well. Improving outcomes by delaying operation may serve to "select out" hardier patients, but other patients who might benefit from surgical intervention will die.

	Footnotes

Top Footnotes Abstract Introduction Material and methods Results Comment References

 
This article has been selected for the open discussion forum on the STS Web site: http://www.sts.org/annals

	References

Top Footnotes Abstract Introduction Material and methods Results Comment References

 

1. Bisno A.L., Dismukes W.E., Durack D.T., et al. Antimicrobial treatment of infective endocarditis due to viridans streptococci, enterococci, and staphylococci. JAMA 1989;261:1471-1477.[Medline]
2. Pruitt A.A., Rubin R.H., Karchmer A.W., Duncan G.W. Neurologic complications of bacterial endocarditis. Medicine (Baltimore) 1978;57:329-343.[Medline]
3. Middlemost S., Wisenbaugh T., Meyerowitz C., et al. A case for early surgery in left-sided endocarditis complicated by heart failure: results in 203 patients. J Am Coll Cardiol 1991;18:663-667.[Abstract]
4. Kimose H.H., Lund O., Kromann-Hansen O. Risk factors for early and late outcome after surgical treatment of native infective endocarditis. Scand J Thorac Cardiovasc Surg 1990;24:111-120.[Medline]
5. Eishi K., Kawazoe K., Kuriyama Y., et al. Surgical management of infective endocarditis associated with cerebral complications. J Thorac Cardiovasc Surg 1995;110:1745-1755.[Abstract/Free Full Text]
6. Ting W., Silverman N., Levitsky S. Valve replacement in patients with endocarditis and cerebral septic emboli. Ann Thorac Surg 1991;51:18-22.[Abstract]
7. Gillinov A.M., Shah R.V., Curtis W.E., et al. Valve replacement in patients with endocarditis and acute neurologic deficit. Ann Thorac Surg 1996;61:1125-1130.[Abstract/Free Full Text]
8. Von Reyn C.F., Levy B.S., Arbeit R.D., et al. Infective endocarditis: an analysis based on strict case definitions. Ann Intern Med 1981;94(4 Pt 1):505-518.
9. Gentry L.O., Khoshdel A. New approaches to the diagnosis and treatment of infective endocarditis. Tex Heart Inst J 1989;16:250-257.
10. Aslamaci S., Dimitri W.R., Williams B.T. Operative considerations in active native valve infective endocarditis. J Cardiovasc Surg (Torino) 1989;30:328-333.[Medline]
11. Cowgill L.D., Addonizio V.P., Hopeman A.R., Harken A.H. A practical approach to prosthetic valve endocarditis. Ann Thorac Surg 1987;43:450-457.[Abstract]
12. Jones H.R., Siekert R.G. Neurological manifestations of infective endocarditis. Stroke 1989;112:1295-1315.
13. Garvey G.J., Neu H.C. Infective endocarditis—an evolving disease: a review of endocarditis at the Columbia-Presbyterian Medical Center. Medicine (Baltimore) 1978;57:105-127.[Medline]
14. Hart R.G., Kagan-Hallet K., Joerns S.E. Mechanisms of intracranial hemorrhage in infective endocarditis. Stroke 1987;18:1048-1056.[Abstract/Free Full Text]
15. Hart R.G., Foster J.W., Luther M.F., et al. Stroke in infective endocarditis. Stroke 1990;21:695-700.[Abstract/Free Full Text]
16. Salgado A.V., Furlan A.J., Keys T.F., Nichols T.R., Beck G.J. Neurologic complications of endocarditis: a 12-year experience. Neurology 1989;39:173-178.[Abstract/Free Full Text]
17. Le Cam B., Guivarch G., Boles J.M., et al. Neurologic complications in a group of 86 bacterial endocarditis patients. Eur Heart J 1984;5(Suppl C):97-100.

This article has been cited by other articles:

				Task Force Members, D. Horstkotte, F. Follath, E. Gutschik, M. Lengyel, A. Oto, A. Pavie, J. Soler-Soler, G. Thiene, A. von Graevenitz, et al. Guidelines on Prevention, Diagnosis and Treatment of Infective Endocarditis Executive Summary: The Task Force on Infective Endocarditis of the European Society of Cardiology Eur. Heart J., February 1, 2004; 25(3): 267 - 276. [Full Text] [PDF]

				C. Stollberger, J. Finsterer, A. Pratter, W. Kopsa, J. Preiser, and A. Valentin Ischemic Stroke and Splenic Rupture in a Case of Streptococcus bovis Endocarditis J. Clin. Microbiol., June 1, 2003; 41(6): 2654 - 2658. [Abstract] [Full Text] [PDF]

				E. Mylonakis and S. B. Calderwood Infective Endocarditis in Adults N. Engl. J. Med., November 1, 2001; 345(18): 1318 - 1330. [Full Text] [PDF]

				M. Heiro, J. Nikoskelainen, E. Engblom, E. Kotilainen, R. Marttila, and P. Kotilainen Neurologic Manifestations of Infective Endocarditis: A 17-Year Experience in a Teaching Hospital in Finland Arch Intern Med, October 9, 2000; 160(18): 2781 - 2787. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Irving L. Kron Curtis G. Tribble Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Parrino, P. E. Articles by Tribble, C. G. Search for Related Content PubMed PubMed Citation Articles by Parrino, P. E. Articles by Tribble, C. G.