Extracranial ectasia and embolic infarcts in HIV: Two Case Reports and A Clinical Decision-Making Algorithm.

Abstract

HIV is known to increase the risk of both ischemic and hemorrhagic stroke. There are many postulated mechanisms for this elevated risk including an HIV-induced vasculopathy and/or coagulopathy, opportunistic infections, and cardioembolic etiologies, among others. Regarding vasculopathy, prior reports have described the various changes to the arterial vasculature that can occur in the setting of HIV, yet the appropriate work-up and management of this condition remains poorly defined. Here we describe two cases of patients with HIV presenting with large vessel intracranial occlusions in the setting of ectatic extracranial vasculature accompanied intraluminal thrombus formation. One patient underwent thrombectomy, while the other improved after receiving IV-tPA. Inferring on these cases and the existing literature, a standardized work-up and treatment algorithm is proposed, emphasizing the key management decisions that should be considered on a case-by-case basis.

Introduction:

Numerous studies have demonstrated that patients with HIV have a higher prevalence of both ischemic and hemorrhagic stroke (1–4). While the precise risk mechanisms may be stroke subtype specific, atherogenesis and endothelial disruption are consistently implicated (5). A few case reports describe embolic appearing infarcts in the setting of HIV and large vessel extracranial atherosclerosis, inferring a symptomatic atherosclerotic mechanism and/or an aberrant coagulation process (6). Other studies have demonstrated arterial remodeling in the setting of HIV resulting in dolichoectasia (7). Certainly, these various vascular changes could disrupt cerebral autoregulation, diminish vessel wall tensile strength, and disrupt endothelial function, thereby resulting in an increased risk of stroke. Here we describe two patients with HIV that had evidence of large extracranial artery ectasia with subsequent thrombus formation leading to embolic infarcts. Using these cases and reviewing the existing literature, we propose a workup and treatment algorithm emphasizing the key diagnostic and treatment considerations in this setting.

Case presentations:

Case 1

A 56-year-old male with a past medical history of HIV on Triumeq (CD4 625, viral load 54), hypertension, active cocaine and tobacco use as well as a prior left MCA territory stroke three years earlier (at that time received IV-tPA with no residual deficit) presented to an outside hospital with right hemiparesis and aphasia with NIHSS of 17. Per history he had used cocaine a few hours prior to presentation. CTA showed a distal left M1 occlusion and left fetal PCA. The patient presented outside of the IV-tPA window and was subsequently transferred to our Comprehensive Stroke Center where an MRI demonstrated encephalomalacia consistent with the known prior L MCA stroke and acute stroke as per restricted diffusion involving the deep left MCA territory and a small focus in the occipital lobe (Figure 1A). Diffusely ectatic common carotid arteries were noted bilaterally (Figure 1B). The patient underwent thrombectomy approximately 17 hours after symptom onset with a resultant grade 3 thrombolysis in cerebral infarction (TICI) score.

Figure 1:

(A) Diffusion weighted imaging depicting acute stroke in the left insula, basal ganglia and occipital region. (B) Ectatic common and internal carotid arteries bilaterally. (C) Plaque vs. non-occlusive thrombus at origin of left internal carotid artery.

The patient’s stroke work-up included a TTE demonstrating mild left ventricular hypertrophy and a moderately dilated aortic root, an LDL of 64, and an HgbA1C of 5.1. On further review of his imaging with neuroradiology, he was noted to have a chronic plaque versus non-occlusive thrombus in the left ICA (see Figure 1C). Interestingly this was also noted on imaging during his prior stoke-workup from 3 years prior, read as “nonatheromatous plaque within the left proximal ICA with 10% luminal stenosis.” In relation to his current stroke, this stenosis was thought to be the likely source of embolization, which was further corroborated by a transcranial doppler emboli detection study demonstrating 83 left-sided high intensity transient signals over a 30-minute evaluation period. Given these findings, the patient was started on dual antiplatelet therapy for 90 days as based on interpretation of the SAMMPRIS trial (8). He was discharged to rehab where he developed a DVT and was subsequently switched from dual antiplatelet to anticoagulation.

Case 2:

A 45-year-old female with past medical history of HIV non-compliant with antiretroviral therapy, previously on Genvoya (CD4 202, viral load 48,879) and tobacco use presented to an outside hospital with left hemiparesis and NIHSS of 11. The patient was reported to have been using K2 a few hours prior to presentation. Head and neck CTA showed thrombus in the right ICA and M1 segment of the MCA. She received IV-tPA at the outside hospital and was transferred to our comprehensive stroke center for thrombectomy. The patient clinically improved by the time of arrival to our facility, with a repeat CTA demonstrating recanalization of the M1 segment with a persistent right M3 and distal ICA occlusion (Figure 2A). MRI demonstrated multiple scattered areas of infarct in the right MCA territory (Figure 2B). The CTA demonstrated ectatic common carotid arteries bilaterally with distal narrowing on the right (Figure 2C).

Figure 2:

(A) CTA showing persistent distal right M3 occlusion on arrival to our facility. (B) Diffusion weighted imaging with scattered areas of infarction within the right MCA territory. (C) Ectatic bilateral common carotid arteries with distal narrowing on the right. (D) Extension of thrombus to petro-cavernous junction of R ICA (see arrow).

On the basis of the imaging the stroke etiology was thought to be HIV vasculopathy or potentially a dissection. A subsequent MRA with fat suppression ruled out dissection but did demonstrate that the ICA thrombus had further propagated forward since the initial imaging, now extending from the right ICA origin to the distal petro-cavernous junction (Figure 2D). Given the clot propagation and risk for further infarcts, the patient was started on anticoagulation utilizing a heparin drip. In the setting of an acute stroke, and to reduce the risk of hemorrhagic conversion, no heparin bolus was given with weight-based PTT goal of 50–70. Additional stroke work-up included HgbA1C of 5.6, LDL of 113, TSH of 2.6 and negative hypercoagulable laboratories including antithrombin III, cardiolipin antibody, beta 2 glycoprotein, factor V Leiden, and lupus anticoagulant. Despite anticoagulation, repeat vessel imaging a few days later again demonstrated a persistent right ICA thrombus, which had now propagated into the cavernous portion of the ICA. The patient remains stable clinically and was subsequently bridged to warfarin. A final CTA obtained prior to discharge demonstrated no further progression of the right ICA thrombus.

Discussion

There is an increased incidence of ischemic stroke in HIV positive patients, often occurring at younger ages in comparison to patients without HIV. Existing literature describes several possible etiologies for these findings including vasculopathy, opportunistic infection, cardioembolism, and coagulopathy (6). Of these, the most common etiologies are thought to be HIV vasculopathy and opportunistic infections (9).

As consistent with our described cases, HIV vasculopathy often involves irregularities of medium to large arteries of the head and neck occurring in the setting of HIV infection. While the precise mechanisms for these changes remain to be fully elucidated, they can be classified on the basis of imaging findings. A study by Benjamin et. al. (9) classified HIV vasculopathy into subsets defined as vasculitis, accelerated atherosclerosis or nonatherosclerotic vasculopathy with intimal hyperplasia; the latter of which can in turn lead to aneurysmal vessel dilation, commonly termed as ectasia. Vessel ectasia can occur with or without atherosclerosis and can eventually lead to either stenosis or vessel occlusion in the setting of acute thrombus formation. Proposed mechanisms for vessel dilation in the setting of HIV include: i) fragmentation of the internal elastic lamina, which may be an early histologic finding; ii) thinning of the media layer (7,10), and; iii) intimal hyperplasia (10,11). Though the specific mechanisms for vessel wall abnormalities remain unclear, it is a definitive contributor to the high prevalence of cerebrovascular disease seen in this population.

The two cases presented both demonstrated evidence of ectatic extracranial vasculature, a finding also often seen intracranially. In a study by Choudry et al. [12], variations in normal common carotid size were quantified demonstrating a mean midpoint common carotid diameter as 6.4 mm+/− 0.7 on the right, 6.5 +/− 0.7 on the left. As depicted in Figures 1B and 2C, the common carotid diameter in both of our presented cases significantly exceeded this proposed mean.Few studies have looked solely at ectatic vasculature, with a heavier focus placed on dolichoectatic (tortuous and dilated) vessels and its role in acute stroke. Gutierrez et. al. described the histologic changes seen in the dolichoectatic vasculature in an HIV population with stroke (10). In this report, arterial changes were subdivided into outward and inward remodeling of vessels as measured by the lumen to wall ratio. The presence of inward remodeling, leading to stenosis and atherosclerosis coincided with hypertension, tobacco use and other vascular risk factors, as would be expected. Outward remodeling on the other hand, was associated with thinning of the media layer and was primarily associated with cryptogenic stroke, representing 21% of the infarcts described in the study (10). Stagnant flow in the dolichoectatic vessel was postulated as the mechanism for thrombus formation that ultimately led to an embolic infarction, potentially consistent with our two cases.

Overall, the appropriate work-up and management of patients with HIV and ischemic stroke remains poorly defined. Beyond the standard “heart to head” stroke work up, Benjamin et al. suggests additional testing for opportunistic infections, including lumbar puncture and chest imaging (9). Based on our described cases, and other data implicating intrinsic vascular changes as a stroke mechanism, thorough arterial imaging is a must to infer on stroke etiology in such situations. Given the lack of literature regarding the management of patients with ischemic stroke and HIV vasculopathy, we reviewed literature on intraarterial dolichoectasia in the non-HIV population and here propose a workup and treatment algorithm to help guide other physicians. Although we do emphasize that treatment and management decisions must be made on a case-by-case basis.

As described in literature regarding non-HIV populations, the pathophysiology of intraarterial dolichoectasia (IADE) is thought to be due to impairment in matrix metalloproteinases (MMPs) (13,14). Genetically, the MMP3 5A genotype has been associated with abdominal-aortic and coronary artery aneurysms (14,15). IADE is more frequently seen in the posterior circulation (14). In general, and contrary to expectations, IADE has been found to be associated with manifestations of small vessel disease, including lacunar infarcts, leukoaraiosis and cerebral microbleeds (16–18). While studies differed in how dolichoectasia was radiologically evaluated, the majority utilize high resolution MRA with gadolinium, which is thought to be more sensitive than CTA (13,14).

While ectasia and dolichoectasia can lead to ischemic stroke through thrombotic and embolic mechanisms, other intracranial consequences can include mass effect on surrounding structures (especially on cranial nerves and the brainstem in vertebrobasilar dolichoectasia), compression of the third ventricle leading to obstructed CSF flow, and vessel-wall rupture leading to intracranial hemorrhage. Given the high risk for these outcomes, Pico et al. (13) suggested an algorithm for evaluation of IADE, again recommending an initial evaluation with MRI and a high-resolution gadolinium enhanced MRA in patients presenting with any of the above-described conditions. If there is presence of dolichoectasia, patients should be evaluated for vascular risk factors, evidence of small vessel disease and abdominal aortic aneurysm. Younger patients without vascular risk factors should additionally be screened for metabolic diseases (e.g. Fabry, Pompe), infectious (i.e. HIV) and other genetic causes. Based on this algorithm, if dolichoectasia is symptomatic or if it is incidentally found on imaging, patients should undergo serial imaging at 6 months and then yearly to evaluate for vessel stability versus enlargement. As appropriate, surgery or endovascular therapy can be considered.

Notably, there have been no controlled trials to evaluate optimal secondary prevention measures in the setting of ischemic stroke as related to ectatic or dolichoectatic vasculature. Studies have shown an association between the pathophysiology of dolichoectasia and small vessel disease (19), supporting the premise that aggressive management of vascular risk factors is a key factor in these patients. Inferring from the aneurysm literature, smoking cessation is critical as well (20). Another important consideration in secondary prevention is the use of anti-platelet agents versus anticoagulation. Again, in this setting there are no controlled trials to compare these two treatments, neither over short- or long-term time periods. Overall, there is a preference towards antiplatelets given the higher risk of intracerebral hemorrhage in patients with intracranial dolichoectasia (13, 14).

Based upon the described studies, and information available for the management and treatment of non-HIV dolichoectasia, we propose a stroke management algorithm when HIV-related intracranial or extracranial ectasia is identified (Figure 3). This emphasizes the importance of considering more prevalent and alternate stroke etiologies in this population as a part of the diagnostic work up. In the end, optimal control of all vascular risk factors over the long-term is paramount.

Figure 3:

Proposed algorithm for inpatient management of ischemic stroke in HIV.

First, as consistent with other stroke patients, all HIV patients with stroke or suspected TIA warrant an expedited evaluation that can be simply defined as from ‘heart to head’. In other words, evaluations of the heart, the proximal aorta, the vasculature of the head and neck, as well as clinical and laboratory testing related to vascular risk factors should be performed on an inpatient basis. If the patient is presenting with stroke symptoms within 24 hours from time-of-symptom-onset, they should undergo an emergent non-contrast brain computed tomography (CT) brain, a multi-phase CT angiogram (CTA), and as based upon timing and results they may be eligible for IV-tPA, thrombectomy, or both. Retrospective studies have shown similar risks of death with thrombolysis in HIV and non-HIV infected patients (21). Further work up includes a transthoracic echocardiogram (TTE) with a bubble study to evaluate for a potential cardioembolic source [22]. Finally, specific to this population, it is important to assess CD4 count and viral load to determine the risk of concurrent opportunistic infection leading to stroke. A broad differential should be considered in all HIV patients, emphasizing opportunistic infections.

If the initial imaging is concerning for a vasculitis, black-blood MRI can be used to evaluate for active inflammation, which may also infer on hemorrhage risk. If HIV vasculopathy is suspected or identified, this can be further evaluated with high resolution MRA with gadolinium enhancement. In these cases, a transcranial doppler emboli detection study can also be considered to determine embolic potential which may infer on the use antiplatelet versus anticoagulation. Additionally, if there is evidence of ectasia, one should consider serial imaging to evaluate for stability over time. We suggest follow-up imaging in 3–6 months post-presentation and then annually to confirm stability. Aggressive risk factor optimization is emphasized, especially smoking cessation, which has been shown to contribute to aneurysm development and small vessel disease (20). Given bleeding risks, antiplatelet therapy is generally preferred over anticoagulation, although short-term anticoagulation can be considered in some situations, as consistent with our described cases. While combined antiretroviral treatment has been shown to decrease all-cause mortality in patients with HIV, its role in stroke reduction is unclear, with some evidence pointing towards an increased stroke risk with certain antiretrovirals (6). Nonetheless, optimal and long-term control of HIV is a key component to preventing the potential downstream HIV effects, including ectasia.

Conclusion

There are many possible etiologies for stroke in patients with HIV. While HIV vasculopathy is a leading contributor, the mechanisms by which this condition, particularly how extracranial ectasia increases ischemic stroke risk, remain poorly understood. The presented management algorithm provides a basic framework for clinicians to consider when evaluating such patients, emphasizing potential stroke etiologies, workup and treatment options.