Premature vascular disease in young adult stroke: a pathology-based case series.

Abstract

Introduction:

The prevalence of modifiable vascular risk factors is increasing in young adults and may contribute to the growing frequency of stroke in this population. The neuropathology and endorgan damage profile of young adult stroke patients with clinically advanced atherosclerosis or arteriosclerosis has not been studied.

Methods:

This retrospective study included patients aged 18–60 years admitted to our hospital from 1995–2017 with recurrent ischemic or hemorrhagic strokes, fatal stroke, or stroke associated with advanced small vessel disease (SVD) on brain MRI, who had no evidence for structural, genetic, inflammatory or infectious etiology for stroke, and had adequate pathological materials available for analysis. The presence of atherosclerosis, arteriolosclerosis, left ventricular hypertrophy and nephrosclerosis were evaluated.

Results:

Twelve patients (mean age 47±9 years, range 31–57 years, 67% male) met inclusion criteria. Four had fatal intracerebral hemorrhage (ICH), 3 had recurrent non-fatal ICH, one had ICH with advanced SVD on MRI, and 4 had recurrent ischemic strokes including two with transient ischemic attacks. Pathological studies showed moderate/severe atherosclerosis in 64% and moderate/severe arteriolosclerosis in 41% of patients. Pathological data to evaluate end-organ damage were available for nine patients; 8 showed left ventricular hypertrophy and all showed nephrosclerosis.

Conclusion:

Young adult stroke patients with recurrent stroke, fatal stroke, or SVD on imaging have advanced atherosclerosis and arteriolosclerosis-related pathological changes in multiple organ systems. Aggressive control of atherosclerosis risk factors is warranted even in young individuals.

INTRODUCTION

Recent studies have shown an increasing rate of stroke hospitalizations in young adults over the last two decades, potentially attributable to better detection as well as a concomitant increase in modifiable vascular risk-factors such as hypertension, dyslipidemia, obesity, physical inactivity, increased illicit drug use, and smoking.^1–5 Young stroke patients demonstrate a considerable burden of atherosclerosis on vascular imaging, and show significant small vessel disease (SVD) on parenchymal brain MRIs.⁶ Furthermore, young stroke patients have a degree of SVD comparable to patients 10–20 years older, supporting the hypothesis of accelerated cerebral aging and increased susceptibility to vascular risk factors.³

While clinical and imaging evidence of atherosclerosis and cerebral SVD has been described,^3,6 pathology-verified vascular burden including evaluation of intra/extracranial atherosclerosis, arteriolosclerosis, nephrosclerosis and myocardial hypertrophy has been inadequately studied in non-elderly stroke patients. Pathologic confirmation of severe systemic and cerebral vascular disease in young individuals with stroke will provide further evidence to support preventive strategies with aggressive vascular risk-factor control in this population. In this study, we evaluated the pathology of brain and systemic end-organ damage in younger stroke patients with clinical evidence for advanced atherosclerosis and arteriolosclerosis.

METHODS

This retrospective study was approved by the Partners Human Research Committee. The Massachusetts General Hospital Research Patient Database Repository was queried for patients aged 18–60 years admitted to our hospital from 1995–2017 with primary ICD-9 or ICD-10 discharge codes for cerebral infarction or non-traumatic intracerebral hemorrhage who underwent an open brain biopsy or autopsy. Of the 253 patients retrieved, 12 met the following inclusion criteria for stroke from clinically advanced vascular disease and had adequate pathologic materials for evaluation: recurrent ischemic stroke (with/without transient ischemic attack), recurrent spontaneous intracerebral hemorrhage [ICH]), fatal stroke (ischemic or hemorrhagic), or stroke (ischemic or hemorrhagic) associated with advanced SVD on brain MRI. Included patients had no evidence for structural (i.e. dissection, vascular malformations), genetic, inflammatory (e.g. vasculitis) or infectious etiology for stroke on pathological, laboratory and imaging studies. For detailed patient selection see Figure 1.

Figure 1. Patient Selection Flow-chart.

RPDR: Research Patient Database Repository

Clinical data collection

Demographic and clinical data were collected from chart review by stroke neurologists for the following variables: age, gender, race, past medical history of ischemic or hemorrhagic stroke, hypertension, diabetes, dyslipidemia, obesity, atrial fibrillation, myocardial infarction, cirrhosis, alcohol use, and smoking. Laboratory data were reviewed for echocardiography findings (left ventricular hypertrophy), creatinine (mg/dL) and estimated glomerular fraction rate (ml/min). Imaging data were collected from head CT, brain MRI, CT- and MR-angiography, and cerebral angiography. Ischemic stroke mechanism was ascribed using the Causative Classification of Stroke schema.⁷

MRI Markers

Brain MRI markers related to SVD were evaluated based on the STRIVE criteria.⁸ To calculate the total SVD score one point was given for the presence of each of the following features on the total SVD score: a) moderate to severe white matter hyperintensity; b) ≥ 1 lacunar infarct; c) >20 basal ganglia enlarged perivascular spaces; ≥ 1 cerebral microbleeds.⁹ Advanced SVD was defined as a total SVD Score ≥ 2.

Pathological data

After a careful review of all slides from each patient by experienced pathologists (W.S., M.P.F.), 12 proved to have adequate material for evaluation and were included for detailed analysis (see figure 1). All specimens were processed with formalin fixation, standard paraffin embedding, Hematoxylin-Eosin and Luxol Fast Blue. Data was gathered on the severity of arteriolosclerosis(scale 0–3), atherosclerosis (scale 0–3), presence of amyloid angiopathy, severity of white matter rarefaction (scale 0–3), degree of perivascular neuropil dropout (scale 0–3), number of cerebral microinfarcts (none, 1, or ≥2), and microhemorrhages (amount of perivascular hemosiderin; scale 0–3)¹⁰. In autopsy cases, atherosclerosis was assessed grossly at brain cutting. Examined vessels included the internal carotid arteries, basilar artery, vertebral arteries, and circle of Willis. External pictures of the autopsied brains were also reviewed. Histological examination of autopsies specimens included assessment of cortex, subcortical white matter, cerebellum, and basal ganglia. See Figure 2 for pathological features evaluated in this study. In autopsy cases, we retrieved information regarding left ventricular hypertrophy and nephrosclerosis.

Figure 2. Pathological features evaluated in the study.

A-small penetrating artery with severe sclerotic narrowing of the lumen. B-large-sized vessel with an atherosclerotic plaque (*) consisting of old fibrosis and calcifications. C-Luxol-blue staining shows a pink region of white matter loss, which contrasts the deep blue area of retained myelination (upper left). D-multiple vessels display lots of perivascular neuropil (white spaces around the vessels). E-microinfarct showing a small cavity of destroyed neuropil filled with foamy histiocytes. This lesion was not seen at gross. F-small penetrating vessel within the white matter demonstrates scattered hemosiderin-laden macrophages consistent with old microhemorrhages.

Statistical analysis

We compared differences between patients with cerebral ischemic events and those with ICH using Chi Square tests and t-test when appropriate. A value of p<0.05 was considered significant.

RESULTS

Twelve patients (mean age 47±9 years at time of first stroke; 67% male) were included in this pathology-based case series (see Figure 1 for patient selection). Our final study sample comprised a) four patients with fatal ICH; b) four patients with non-fatal ICH, including 3 with recurrent ICHs and one with advanced SVD on brain MRI; c) four patients with recurrent ischemic stroke. Table 1 shows a summary of the demographic, clinical, pathologic and neuroimaging features. Tables 2–4 details the clinical, laboratory and pathologic features of each patient.

Table 1.

Summary of clinical and imaging characteristics

Variables	N=12 (%)
Age, mean±SD (years)	47±9
Sex, female	4 (33)
White	9 (75)
Hypertension	11 (92)
Diabetes	7 (58)
Hyperlipidemia	6 (50)
Atrial Fibrillation	2 (17)
Alcohol use	2 (17)
Smoking	1 (8)
Obesity	4 (33)
Moderate/severe atherosclerosis*	7 (64)
Moderate/severe arteriolosclerosis	5 (42)
Left ventricular hypertrophy†	8 (90)
Glomerulo/nephrosclerosis†	9 (100)
Lacunes‡	7 (77)
Cerebral microbleeds	5 (56)
Moderate/severe WMH‡	7 (77)
BG EPVS > 20‡	5 (55)
SVD score ≥ 2‡	6 (66)

Data expressed in number and (%) except for age

^*Available in 11 patients

^†Available in 9 patients

^‡MRI was available for 9 patients

SD: Standard deviation, WMH: White matter hyperintensities, BG: basal ganglia, CSO: centrum semiovale, EPVS: enlarged perivascular spaces, SVD: small vessel disease.

Table 2.

Detailed Clinical Features

Case	Age* Sex	Race	IS/TIA	CCS	ICH N, type	HTN	Diabetes	HL	Other
1	35, M	Black	0		1, deep	Yes	Yes	No	Cirrhosis
2	52, F	White	1 IS/1 TIA	LAA	1, SDH	Yes	Yes	No	ESRD
3	57, M	White	0		1, deep	Yes	Yes	Yes	None
4	41, M	Hispanic	0		2, deep	No	Yes Type 1	No	Cirrhosis
5	31, M	White	0		2, deep, cerebellar	Yes	No	No	ESRD
6	55, M	White	0		1, deep	Yes	Yes	Yes	ESRD
7	47, F	White	0		1, deep	Yes	Yes	No	None
8	48, F	Asian	0		1, lobar	Yes	No	No	None
9	55, F	White	2 IS	LAA	No	Yes	Yes	Yes	ESRD
10	36, M	White	0		1, deep	Yes	No	Yes	None
11	53, M	White	1 IS/1 TIA	LAA	No	Yes	No	Yes	None
12	54, M	White	Multiple IS	LI	1, deep	Yes	No	Yes	Astrocytoma

^*Age at first ischemic or hemorrhagic event

ESRD, end-stage renal disease; IS: ischemic stroke; TIA: transient ischemic attack; LAA: large artery atherosclerosis; LI: lacunar Infarct; ICH: intracranial hemorrhage; SDH: subdural hematoma; HTN: hypertension; HL: hyperlipidemia.

Table 4.

Detailed Pathologic Findings and Cause of Death

Case No.	Procedure, Age (years)	Arteriolosclerosis	Atherosclerosis	Kidney	LVH	Cause of death
1	Autopsy, 35	Mild	None	NS	Yes	ICH
2	Autopsy, 55	Moderate	Moderate	NS	Yes	trauma, SDH
3	Autopsy, 57	Mild	Severe	NS	Yes	ICH
4	Autopsy, 52	Moderate	Mild	NS	Yes	ICH
5	Autopsy, 58	Severe	Mild	NS	Yes	Ml
6	Autopsy, 61	Mild	Moderate	NS	Yes	Arrhythmia
7	Biopsy, 48	Mild	NA	NA	NA	Alive
8	Surgery, 48	No	Moderate	NA	NA	Alive
9	Autopsy, 55	Mild	Severe	NS	Yes	Ml
10	Autopsy, 36	Moderate	Mild	NA	NA	ICH
11	Autopsy, 58	NO	Severe	NS	No	Metastatic prostate cancer
12	Autopsy, 67	Severe	Moderate	NS	Yes	NA

NS: nephrosclerosis; GS: glomerulosclerosis; LVH: left ventricular hypertrophy on autopsy; ICH: intracranial cerebral hemorrhage; IS: ischemic stroke; NA: non-available; MI: myocardial infarction; In case 10, slides were not available however arteriolosclerosis and atherosclerosis were described in the neuropathology report.

Imaging

On vascular imaging, among 8 patients with ICH, only one who had non-fatal ICH showed evidence for intra/extracranial atheromatous disease (occluded middle cerebral artery and moderate stenosis of the left vertebral artery). In contrast, three of the four patients with ischemic stroke had significant cerebral atheromatous disease (intracranial/extracranial) depicted by neurovascular imaging (Table 3). One patient showed moderate atheromatous disease in the vertebral artery and one patient showed > 70% stenosis of the right internal carotid artery and 50–70% stenosis of the contra-lateral internal carotid artery. Finally, one patient who suffered from an ischemic stroke and a transient ischemic attack showed multiple severe stenosis in the intracranial circulation (right anterior cerebral artery, right middle cerebral associated to an occlusion of the left medial cerebral artery, and basilar artery). Brain MRI showed moderate to advanced SVD in 4 of the 8 patients with any ICH (fatal or non-fatal), and 2 of the 4 patients with ischemic stroke (Table 3).

Table 3.

Detailed Laboratory Features

Case No.	TTE	Infarcts/ICH on CT	Infarcts/ICH on MRI	Extracranial atherosclerosis*	Intracranial atherosclerosis*	SVD score
1	No LVH	Pontine ICH	NA	NA	NA	NA
2	No LVH	SDH	Small Llobar ICH, SDH	VA: R moderate, L mild	Syphons: mild RL	4
3	NA	L BG ICH	NA	Basilar: mild	No	NA
4	NA	L thalamic and small pontine ICH	L thalamic and small pontine ICH	NA	NA	1
5	LVH	L thalamic ICH	L thalamic ICH	NA	NA	2
6	LVH	L pontine ICH	L pontine ICH	None	None	3
7	No	Lthalamic ICH	Lthalamic ICH DWI + lesions	None	None	4
8	No	R PO ICH	R PO ICH DWI + lesions	None	R-MCA: occluded; L-VA: moderate	2
9	LVH	Bilateral CR infarcts	Bilateral watershed infarcts	ICA: R severe, L moderate	None	1
10	LVH	R BG ICH	NA	None	None	NA
11	No	L temporal and parietal lobes chronic infarcts	L temporal and parietal infarcts	None	MCA: L occlusion, R severe; ACA: R severe, L mild; PCA: L and R mild BA: severe	1
12	LVH	Bilateral multiple lacunar infarcts	Chronic L BG ICH, Bilateral multiple lacunar infarcts	None	None	4

^*Presence of stenosis/occlusion on CT/MR Angiography

TTE: transthoracic echocardiogram, LVH: left ventricular hypertrophy; R: right; L: left; NA: non-available; ICH: intracranial hemorrhage; ICA: internal carotid artery; ACA: anterior cerebral artery; PCA: posterior cerebral artery; PO: parieto-occipital; BA: basilar artery; CR: corona radiata; BG: basal ganglia, CSO: centrum semiovale; VA: vertebral artery; DWI: diffusion weighted imaging; MCA: middle cerebral artery; SVD score: total small vessel disease score on MRI;

Pathology

Pathologic material used for analysis was retrieved from full autopsy in 10 patients, from brain biopsy in one patient and hematoma evacuation in one patient. Mean age at pathology acquisition was 52±9 years.

In the fatal ICH group (n=4), 2 cases also had liver cirrhosis as a significant comorbidity. One patient had severe cerebral atherosclerosis on pathology and 2 patients showed moderate arteriolosclerosis. Pathologic material for the evaluation of cerebral lesions and systemic involvement was available for three patients. Multiple cerebral micro-infarcts and microhemorrhages were present in all patients. Advanced white-matter loss and advanced perivascular neuropil dropout were present in 1 patient. Left ventricular hypertrophy and nephrosclerosis were present in all 3 patients.

Four patients had non-fatal ICH, including three with one recurrent non-fatal ICH and one with nonfatal ICH and advanced SVD on MRI. Of these, 2 had hypertensive end-stage renal disease as comorbidity. Two patients had moderate cerebral atherosclerosis and two patients showed moderate/severe arteriolosclerosis on pathology. Cerebral microinfarcts were present in two, and microhemorrhages were present in three non-fatal ICH patients. Advanced stages of white-matter loss and perivascular neuropil dropout were absent in all non-fatal ICH patients. Two non-fatal ICH patients had full autopsy and all showed left ventricular hypertrophy and nephrosclerosis.

Four patients had ischemic stroke and two of them also had one prior transient ischemic attack. Endstage renal disease as a significant comorbidity was present in two patients. In 3 patients, ischemic strokes were secondary to large-artery atherosclerosis and one had exclusively lacunar strokes. Moderate/severe cerebral atherosclerosis was present in all patients and moderate/severe arteriolosclerosis was present in two patients. One patient showed cerebral microinfarcts, two patients showed microhemorrhages, one moderate white matter loss and one severe perivascular neuropil dropout. Three patients showed left ventricular hypertrophy on autopsy. Nephrosclerosis was present in all patients.

Ischemic versus Hemorrhagic Stroke

Finally, we compared clinical, imaging and pathological characteristics between four patients who had an ischemic stroke and 8 patients with an ICH. Clinically, patients with ischemic stroke tended to be more often white (100% vs 63%), have more hyperlipidemia (75% vs 37%), and coronary artery disease (50% vs 13%). The differences in vascular imaging findings are described above. On pathology, ischemic stroke patients had more moderate/severe atherosclerosis (100% vs 42%) and lacunar infarctions (100% vs 37%; all p<0.05). The rates of arteriolosclerosis, diabetes, nephrosclerosis and total SVD score were similar between groups. Patients with ICH tended to have more microinfarcts (63% vs 0%, p>0.05).

DISCUSSION

Our pathology-based case series confirms that brain atherosclerosis and arteriolosclerosis (i.e. large and small vessel disease) are relatively frequent in younger patients with severe symptomatic cerebrovascular disease. We found that systemic involvement is even more frequent, with almost all studied patients showing left ventricular hypertrophy or nephrosclerosis on autopsy. Potentially important differences were identified between ischemic and hemorrhagic stroke. Overall our data suggest that premature vascular risk factor-related vasculopathy can manifest with severe brain and systemic consequences in young or non-elderly individuals. Our results, despite being derived from a relatively small series, have important clinical implications, supporting the idea that strict control of vascular risk factors should start at young ages to prevent the progression of symptomatic cerebrovascular disease.

Our pathology findings showed extensive vascular brain damage, which corroborates the results of recent studies demonstrating a high prevalence of atherosclerosis and cerebral SVD in young stroke patients on neuroimaging.^3,6 We provide data to support the hypothesis of a multi-organ vascular disease, indicated by studies showing association of kidney impairment and MRI markers of cerebral SVD,¹¹ as well as hypertensive myocardial dysfunction and stroke in the young.¹² Although the number of anti-hypertensive medications was not available for most patients, the presence of systemic involvement such as left ventricular hypertrophy and advanced cerebral SVD indicates inadequate blood pressure control.¹³

Due to the limited sample size, we were only able to perform comparisons between ischemic stroke and ICH patients. In this pathologic cohort, ischemic stroke patients tended to be more often white, have more hyperlipidemia, coronary artery disease, more moderate to severe atherosclerosis and lacunes. These findings are in line with known vascular risk factors, comorbidities and pathologic findings that have been extensively related with ischemic vascular disease. We did not find any differences in cerebral arteriolosclerosis between the two subgroups, perhaps due to the limited sample size. However, in young stroke populations MRI markers related to SVD have been reported to be highly prevalent in both ischemic and hemorrhagic population.^14,15 In line with this finding, we found no differences between pathological markers of SVD between ischemic and hemorrhagic stroke patients. Furthermore, MRI SVD scores were similarly high and comparable between both groups. The only pathology marker that was more commonly noted on hemorrhagic stroke patients was cortical microinfarcts. This result could be explained by the high prevalence of diffusion weighted lesions (proposed to represent acute MRI expression of cortical microinfarcts)¹⁶ in ICH population.¹⁷

In our pathologic case series of young stroke patients with severe cerebrovascular disease we found a high prevalence of modifiable risk factors. These findings are supportive of reports that suggest an increasing prevalence of vascular risk factors in recent years coexisting with a higher hospitalization rate for acute ischemic stroke in young adults.¹⁸ Unfortunately, due to the limited sample size we were not able to statistically assess the relationship between the presence of modifiable risk factors and pathologic features of cerebrovascular disease. While recent studies have highlighted the role of genetic risk factors in atherosclerosis, further studies might be needed to identify stroke susceptibility loci specific for accelerated atherosclerosis in young patients.¹⁹

In our study, we also evaluated the presence of severe atheromatous disease evaluated with vascular imaging. As expected, ischemic stroke patients showed a high prevalence of intra/extracranial stenosis. By contrast, only one patient with ICH had moderate cerebral atheromatous disease on vascular imaging. This result is in line with our pathology findings that showed a lower prevalence of moderate/severe atherosclerosis (p<0.05) compared to ischemic stroke patients. However, in ICH patients, pathological assessment demonstrated a two-fold higher increased frequency of moderate/severe cerebral atheromatous than with vascular imaging. This result can be in part explained by the different sensitivity and specificity of both techniques, as neuropathology evaluation is considered the gold standard technique. The difference between vascular imaging and pathology data in hemorrhagic stroke patients might be also explained by a predominant small vessel involvement, not seen in vascular imaging.²⁰ However, this data might also suggest that the relatively high risk of cerebral and extracerebral ischemic disease (mostly related to atheromatous disease) shared by ICH patients on the long term cannot totally be quantified with standard vascular techniques.²¹

We acknowledge an approximately 5-year timespan between stroke event and pathology acquisition; however, we assessed small and large vessel damage at time of stroke by using vascular imaging and MRI. Furthermore, we assume that advanced atherosclerosis and arteriolosclerosis existed already at the time of the event since both pathologic processes would not progress so markedly in this timespan. The definition of “young” is subjective and variable, with the upper age cut-off ranging from 45 years to 55 years in most studies. In our cohort all patients, except one 57year man, were below the age of 55 years at the time of stroke. Unfortunately, we were not able to use any genetic data for analysis.

In conclusion, we found that a widespread and advanced vascular risk factors-related vasculopathy is present in younger stroke patients, resulting in severe cerebral and systemic complications. We identified differences in risk factors, and imaging and pathological findings between ischemic and hemorrhagic stroke patients. Our results suggest that it is never too early for a strict vascular riskfactor control.