Stroke mimics are not uncommon in people with HIV. Of the 98 consecutive patients presenting with an acute focal neurological deficit in Malawi, 11 were reported to have another cause for their presentation (eg, toxoplasmosis, neurocysticercosis, tuberculoma, brain tumour).31
Brain imaging in patients with HIV presenting with sudden-onset focal signs and features suggestive of stroke is therefore essential for diagnosis and to distinguish between cerebral haemorrhage and ischaemic stroke.
Once a diagnosis of stroke is verified, management should be directed towards acute stroke treatment, establishment of the cause of the stroke, management of HIV infection, and secondary prevention of stroke. Concomitant HIV infection raises doubt about extrapolation from evidence-based treatment of non-HIV-related stroke. The role of intravenous thrombolysis is uncertain in the absence of randomised controlled studies in HIV-related stroke. Often, the patient's HIV status is not known at the time of presentation of acute stroke, and the decision of whether to give thrombolysis must be made within a short timeframe. Although no clear evidence of harm exists, and individuals with HIV might well have a stroke that is unrelated to HIV infection, the pathogenesis of stroke can include HIV-associated vasculopathy, infective vasculitis, infective meningitis, and other causes that might increase bleeding risk (panel 1
). Reassuringly, isolated reports of successful use of thrombolysis to treat myocardial infarction in individuals with HIV are available.100
However, the extent to which these findings can be generalised to patients with potentially diseased cerebral vessels and higher bleeding risk with thrombolysis is unclear. Until such data become available, acute therapy, including the use of thrombolysis, will have to be decided on an individual basis, taking into account clinical judgment and patient choice.
Investigation should be directed at assessment for conventional causes of stroke and causes associated with HIV (described above and in ), with particular emphasis on the identification of treatable causes. The approach to management outlined in will not be possible in many areas where HIV infection is prevalent. In low-resource settings, investigation and treatment should be directed at identification of treatable causes of stroke or stroke mimics, such as tuberculosis, cryptococcus, toxoplasmosis, varicella zoster, or herpes virus infection, perhaps by combining CT brain scan, chest radiograph, lumbar puncture (if not contraindicated and in the absence of an alternative cause—eg, an obvious cardioembolic source), and selected blood tests (). Ancillary tests might be needed to establish the cause of stroke (eg, sputum and CSF samples for tuberculosis).101,102
Measuring intrathecal IgG against varicella zoster virus together with CSF DNA PCR improves the likelihood of identifying this potential cause.54
Diagnosis of neurosyphilis in patients with HIV can be complex. A positive CSF venereal disease research laboratory test can help—when this test is negative in a patient with HIV, a test for CSF treponemal antibodies seems a reasonable approach.52,103
Management approach for HIV-infected patients with stroke
Antineutrophil cytoplasmic antibodies (ANCA) assessed by immunofluorescence and enzyme-linked immunosorbent assay have been identified in patients with HIV, but not necessarily in patients with vasculitis, autoimmune disease, or specific opportunistic infection.104–106
Diagnosis of cerebral vasculitis in patients with HIV should be based on the results of appropriate radiological and, if possible, histological features, without the presence of any other potential cause of vasculitis (eg, opportunistic infection). In this setting, detection of ANCA might strengthen the diagnosis.
In high-resource settings, investigation should include detailed assessment of cerebral arteries with carotid doppler and CT, or magnetic resonance angiography, depending on local skills. In selected patients, perhaps those who have had further events or have a comorbidity suggesting autoimmune or other disease, conventional angiography or brain biopsy might be indicated. In low-resource settings without access to brain imaging, we advise basic stroke care with the approach outlined in the South African stroke guideline.107
The role of immunosuppression with corticosteroids is far from clear.32,36
In the absence of any evidence to guide management, it seems reasonable to introduce cART (with a corticosteroid if the patient has a poor response) if vasculitic HIV-associated vasculopathy is suspected and other potential autoimmune or infectious causes have been excluded.
Evidence has convincingly shown that cART results in a reduction of all-cause mortality in patients with HIV.108,109
Far less certain is whether cART treatment, particularly exposure to protease inhibitors, increases the long-term risk of stroke and myocardial infarction as a result of metabolic effects (eg, hypercholesterolaemia, already described) and extended survival (ageing is a risk factor for stroke and some populations infected with HIV have a high prevalence of cigarette smoking).109–111
The risk–benefit ratio of cART based on current knowledge seems to be favourable. However, in view of the concern about long-term stroke and cardiovascular disease risk, a pragmatic approach seems reasonable—ie, physicians should identify and manage risk factors, perhaps change the class of cART regimen, or consider a cholesterol-lowering drug if appropriate.108,109,112–114
None of the studies that guide the use of secondary prevention for stroke, including use of antiplatelets, statins, and blood-pressure-lowering therapy, can be directly extrapolated to patients with HIV who have had a stroke. However, general lifestyle factors and reduction of vascular risk factors seems sensible. Finally, the mode of HIV infection relevant to the patient should be considered, because this might affect underlying stroke risk factors, cause, and management. In sub-Saharan Africa, the major mode of HIV transmission is sexual intercourse among heterosexuals, whereas transmission via IDU is rare.115,116
However, a history of IDU use is relevant, particuarly in areas where it is common, because it might be associated with several potential causes of stroke, including the use of specific drugs (cocaine, amphetamines, sympathomimetic drugs), infective endocarditis, and embolisation of particulate matter.117
In many regions, cigarette smoking is more common in people with HIV than in the general population.118