Strokes in Patients With Injection Drug Use and Tricuspid Valve Endocarditis – A Case Series

Abstract

Research Design: In this study, we describe patients from a tertiary care safety-net hospital endocarditis registry with tricuspid valve infective endocarditis (TVIE), and concomitant acute or subacute ischemic stroke predominantly associated with injection drug use (IDU). We retrospectively obtained data pertinent to neurologic examinations, history of injection drug use (IDU), blood cultures, transthoracic/transesophageal echocardiography (TTE/TEE), neuroimaging, and Modified Rankin Scale (mRS) scores at discharge. Only those patients with bacteremia, tricuspid valve vegetations, and neuroimaging consistent with acute to subacute ischemic infarction and microhemorrhages in two cases were included in this series. Results: Of 188 patients in the registry, 66 patients had TVIE and 10 of these were complicated by ischemic stroke. Neurologic symptoms were largely non-specific, eight patients had altered mental status and only 3 had focal deficits. Nine cases were associated with IDU. Two patients had evidence of a patent foramen ovale on echocardiography. Blood cultures grew S. aureus species in 9 of the patients, all associated with IDU. Three patients died during hospitalization. The mRS score at discharge for survivors ranged 0-4. Conclusions: Patients with strokes from TVIE had heterogeneous presentations and putative mechanisms. We noted that robust neuroimaging is lacking for patients with TVIE from IDU and that such patients may benefit from neuroimaging as a screen for strokes to assist peri-operative management. Further inquiry is needed to elucidate stroke mechanisms in these patients.

Introduction

In parallel with the opioid epidemic, the incidence of infective endocarditis (IE) has doubled over the past 2 decades. ¹ Neurologic complications in IE are frequent in left-sided valvular disease due to septic emboli, whereas stroke due to right-sided IE is uncommon.^2,3 Here, we describe the presentations, neuroimaging, and outcomes of patients with TVIE and ischemic stroke.

Methods

The study protocol was approved by the Institutional Review Board of Boston University School of Medicine. Cases were identified through retrospective review of the Boston Medical Center Endocarditis Inpatient Registry to identify patients with tricuspid valve disease who presented between January 2017 and July 2021. Since January 2017, a multi-disciplinary team including cardiology, cardio-thoracic surgery, addiction medicine, infectious disease, and neurology have followed all patients who were consulted after admission to our hospital with IE. The electronic medical records of these patients were reviewed to abstract demographics, neurologic examinations, history of IDU, blood culture data, echocardiography, and neuroimaging. Patients were included in this case series if they had positive blood cultures, tricuspid valve involvement confirmed on echocardiography (TTE and TEE respectively), as well as neuroimaging by computed tomography (CT) and/or magnetic resonance imaging (MRI) demonstrating evidence of acute and subacute ischemic stroke. Outcomes were assessed based on modified Rankin scale (mRS) scores at discharge and at follow-up when available.

Results

Demographics

Sixty-six from a total of 188 patients (35%) with TVIE were identified in the study period. Ten patients (6 female, 4 male) of diverse racial backgrounds (5 White, 3 Hispanic, 2 Black) were found to have TVIE complicated by septic emboli to the brain (Table 1). The median age of these patients was 38 (29-50) years. All but 1 case was associated with IDU. In this case, the patient’s risk for IE was attributed to central line access during a prior hospitalization.

Table 1.

Patients with TVIE and Stroke.

Age	Sex	Race	Affected valves	Organism	IDU	Neuroimaging	Cardiac imaging (TTE and TEE)	Intracardiac shunt	Neurologic Symptoms and Exam	mRS at discharge	mRS at follow up	HF	Conduction anomaly	Valve replacement
42	M	White	Tricuspid	MSSA	Yes	MRI without contrast: multiple FLAIR hyperintensities of varying size and morphology suggestive of septic emboli. Small focus of susceptibility artifact adjacent to one lesion.	Severe TR, Tricuspid with vegetation and perforation	No	No focal deficits, NIHSS 0	0	No Neurology follow up	No	No	No
50	M	Black	Tricuspid	Enterococcus faecalis	No	MRI brain without contrast: few punctate diffusion weighted hyperintensities consistent with recent embolic infarcts.	Severe TR, Tricuspid with vegetation and possible perforation	No	AMS with NCSE. NIHSS 0. Suspected meningo-encephalitis based on CSF studies.	3	No Neurology follow up	No	No	No
34	F	White	Tricuspid	MSSA, MRSA, Klebsiella species.	Yes	MRI brain/MRA head and neck: multiple small foci of restricted diffusion seen throughout cerebral hemispheres and brainstem.	Severe TR, Tricuspid with vegetation	No	AMS, no NIHSS documented.	0	No Neurology follow up	Subsequent NYHA 1	No	No
38	F	Hispanic	Tricuspid	MRSA	Yes	MRI brain with and without contrast: two small foci of hyper intense FLAIR signal in bilateral frontal white matter.	Severe TR, Tricuspid with vegetation flail posterior leaflet, positive bubble study on TTE, no shunt seen on TEE	Undetermined	AMS, no NIHSS documented.	0	No Neurology follow up	No	No	No
37	F	Black	Tricuspid	MRSA, Candida albicans, Candida glabrata, Candida lipolytica	Yes	MRI brain without contrast: scattered punctate lesions, some demonstrating DWI hyperintensity, some demonstrating punctate enhancement	Severe TR, Tricuspid with vegetation	PFO	AMS, no NIHSS documented.	0	No Neurology follow up	No	No	No
44	M	Hispanic	Tricuspid	MSSA	Yes	MRI brain with and without contrast: multiple periventricular and subcortical T2 and FLAIR hyperintensities. Few demonstrated susceptibilities on SWI and hyperintensities on diffusion.	Severe TR, Tricuspid with vegetation, PFO on TEE	PFO	Loss of consciousness. Weakness in all extremities. Vision changes. AMS. NIHSS 21.	3	1	Subsequent NYHA class III	No	Awaiting
29	F	Hispanic	Tricuspid, Aortic	MSSA	Yes	CTA head and neck: small focus of hypodensity of the right high frontal lobe	Moderate TR, Tricuspid and Aortic with vegetations	No	AMS, comatose and sedated. NIHSS not documented.	6	Deceased	No	Incomplete right bundle branch block	No
35	M	White	Tricuspid, Mitral	MSSA	Yes	MRI brain with and without contrast: scattered distribution of restricted diffusion and T2/FLAIR hyper intensity prominent at the gray white junction, right thalamus, and left splenium of the corpus callosum.	TR, MR, numerous Tricuspid vegetations, 1 vegetation on Mitral	No	Left leg weakness. NIHSS not documented.	6	Deceased	Acute on chronic right heart failure due to valvular disease	No	No
38	F	White	Tricuspid, Mitral	MSSA	Yes	CT head noncontrast: recent infarct in the left basal ganglia with mild amount of bilateral subarachnoid hemorrhage. CTA head and neck: nonocclusive basilar tip thrombus. MRI brain without contrast: multifocal regions of restricted diffusion and T2/FLAIR hyper intensity suspicious for recent infarcts with multifocal regions of blooming on GRE.	TR, MR, Tricuspid and Mitral valves with vegetations and perforations	No	Loss of consciousness, right sided weakness. NIHSS 29.	4	4	Subsequent NYHA III	No	Y – Mitral valve
31	F	White	Tricuspid, Aortic	MRSA	Yes	CT head noncontrast: A 6 mm focus of hypodensity in the left centrum semiovale, which may reflect white matter changes or a septic embolus/subacute infarction	Severe TR, right coronary cusp flail of the aortic valve. Tricuspid and Aortic valve with vegetations	N/A	AMS	6	Deceased	N/A	N/A	No

Abbreviations: AMS: Altered Mental Status; CT: Computed Tomography; CTA: Computed Tomography Angiography; DWI: Diffusion Weighted Imaging; FLAIR: Fluid-Attenuated Inversion Recovery; GRE: Gradient Echo sequences; IDU: Injection Drug Use; MSSA: Methicillin Sensitive Staphylococcus Aureus; MR: Mitral Regurgitation; MRI: Magnetic Resonance Imaging; mRS: Modified Rankin Scale; MRSA: Methicillin Resistant Staphylococcus Aureus; NCSE: Non Convulsive Status Epilepticus; NIHSS: National Institutes of Health Stroke Scale; PFO: Patent Foramen Ovale; SWI: Susceptibility Weighted Imaging; TEE: TransEsophageal Echocardiogram; TR: Tricuspid Regurgitation; TTE: TransThoracic Echocardiogram.

Neurologic Presentation

The neurologic symptoms of these patients were mostly non-specific. Eight had altered mental states ranging from confusion to coma with brainstem dysfunction. Three demonstrated focal deficits, 2 with lateralized weakness, one with vision loss. One patient had non-convulsive status epilepticus and inflammatory cerebrospinal fluid suggesting meningoencephalitis.

Microbiology

All patients had blood culture-proven bacteremia. The one patient without IDU had a central line-associated infection with Enterococcus faecalis. All other infections were caused by Staphylococcus aureus associated with IDU. Two were polymicrobial: 1 with methicillin-sensitive Staphylococcus aureus (MSSA), methicillin-resistant Staphylococcus aureus (MRSA), and Klebsiella co-infection and 1 with MRSA and Candida albicans co-infection. The 7 remaining Staphylococcal cases were monomicrobial (5 MSSA, 2 MRSA).

Echocardiography

All patients underwent transthoracic echocardiography (TTE), and 7 also had transesophageal echocardiography (TEE). Four patients had bilateral vegetations, 2 with concomitant mitral valve and 2 with concomitant aortic valve involvement. Six patients had isolated tricuspid valve involvement. Of these 6 patients, two had evidence of a patent foramen ovale (PFO), and 1 had discordant findings with a positive bubble study on TTE but no PFO on TEE. Three patients with isolated tricuspid valve IE without evidence of intracardiac shunting.

Neuroimaging

All patients had neuroimaging (CT 10, MRI 7) consistent with embolic infarction. Two patients with isolated tricuspid valve involvement had evidence of cerebral microhemorrhage on the susceptibility-weighted MRI sequence. One patient with IE affecting both the tricuspid and mitral valves had evidence of a basilar tip thrombus as well as subarachnoid hemorrhage. Non-invasive vessel imaging did not suggest vasculitis or infectious intracranial aneurysms in any patient. Four patients underwent digital subtraction angiography (DSA) which did show an aneurysm in one case.

Outcomes

Three of the 10 patients died (from coma, severe heart failure, and, sepsis) during hospitalization, all 3 had concomitant left-sided valvular lesions. Survival and functionality was better in those with isolated TVIE. Modified Rankin Scale score (mRS) at discharge was 0 (no disability) in 4 of these patients, 3 (moderate disability) in 2 patients and 4 (moderate severe disability) in the only survivor with bilateral valvular involvement. Three of the survivors developed heart failure and two were further evaluated for valvular replacement. The last patient with isolated TVIE survived without symptomatic cardiac disability. An entire summary of our findings can be found in Table 1.

Discussion

We describe ten cases of TVIE complicated by ischemic stroke, finding that concurrent left-sided valvular infections and intracardiac shunts most commonly explain the etiology of stroke in TVIE. In a minority of patients, the mechanisms of stroke remain elusive. One patient had meningitis which may also contribute to cerebral vasculopathy and infarction, but this was not seen in majority of the cases. Pulmonary shunting, imperfect sensitivity of echocardiography, small vessel vasculitis, paroxysmal atrial fibrillation as mechanisms require further exploration.^4-6 Though not utilized in our series, this limitation to detect extracardiac shunts may be improved upon using transcranial Doppler studies to detect emboli and CTA of the chest to assess for pulmonary shunts. We find it striking that the patients in our series are typically younger and with fewer traditional cardiovascular risk factors compared to typical populations of patients with ischemic strokes.

The neurologic presentation in these cases was largely altered mentation without focal neurologic deficits. We argue that such presentations make stroke identification even more challenging as patients with IDU-associated IE often have alternative explanations for encephalopathy such as acute intoxication, withdrawal, sepsis, or seizure. We suspect embolic shower as the predominant explanation of acute encephalopathy though it is possible that we underdiagnosed seizures/status epilepticus in this cohort. Only three of these patients underwent monitoring with electroencephalography (EEG). One patient had long runs of generalized rhythmic delta activity with 1.5 hertz spike waves that terminated after lorazepam administration suggestive of non-convulsive status epilepticus. EEG recordings on the other two demonstrated diffuse slowing without any epileptiform activity, suggestive of encephalopathy. We acknowledge the limitation of using neuroimaging for the diagnosis of stroke in cases of pure encephalopathy as seizures/status epilepticus may also appear as diffusion restriction on MRI.

Only 30% of the patients in our registry with TVIE who had any neuroimaging had evidence of recent strokes, and only 22% of those had MRIs. As such, patients in our registry with silent cerebral infarcts may be underrepresented. Identification of cerebrovascular complications of IE is crucial in decision-making surrounding anti-thrombotic decisions, candidacy and timing for valvular surgery, and determining prognosis. ² IE-associated stroke can also impact expected patient mortality. Right-sided endocarditis is typically associated with lower mortality than left-sided disease. ⁷ However, patients with TVIE and stroke may be an exception to this generalization.

Given the estimated frequency of PFO in 20% of the population, ⁴ with increased IDU, we may anticipate increased incidence of stroke. Additionally, a minority of patients with TVIE may be at risk for stroke, even in the absence of concomitant left-sided valvular infections or intra-cardiac shunt, for which further studies are required for better understanding of underlying mechanisms. We encourage a wider use of neuroimaging, ideally MRI with vessel imaging, and early involvement of neurologists to assess for strokes in this population.

ORCID iDs

Karan S. Hingorani https://orcid.org/0000-0002-6756-608X

Pria Anand https://orcid.org/0000-0002-7421-7750

Anna M. Cervantes-Arslanian https://orcid.org/0000-0002-6779-0647