Article Text

Download PDFPDF
Successful treatment of HIV associated cerebral vasculopathy with HAART
  1. N J Cutfield1,
  2. H Steele2,
  3. T Wilhelm3,
  4. M W Weatherall4
  1. 1
    Neuro-otology, Charing Cross Hospital, Imperial College London, London, UK
  2. 2
    Newcastle General Hospital, Newcastle, UK
  3. 3
    Royal Free Hospital, London, UK
  4. 4
    Charing Cross Hospital, Imperial College NHS Trust, London, UK
  1. Dr N J Cutfield, Neuro-otology, Imperial College Level 10, Laboratory Building, Charing Cross Hospital, Fulham Palace Road, London W6 8RF, UK; n.cutfield{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

In response to whether an HIV associated vasculopathy may be reversible,1 we describe a man with severe cognitive impairment due to an isolated HIV associated cerebral vasculopathy, likely to be a vasculitis. With initiation of highly active antiretroviral treatment (HAART), and without additional antimicrobial or immunosuppressive treatment, he made a major recovery over 6 weeks to independent living in the community.

A 49-year-old man from Zimbabwe living in the UK for 5 years presented after an unwitnessed collapse and 2 weeks of cognitive decline. He had received a course of oral corticosteroids for parotitis 2 months earlier. He had no other medical history and no conventional vascular risk factors. On examination he was encephalopathic with severe global cognitive impairment confirmed on formal psychometry. He knew his name, but all responses on the Abbreviated Mental Test Score were incorrect. Speech was very slow and he could follow some one step commands. There was right hemi-neglect and he could not stand unsupported. There were no additional neurological signs in the limbs allowing for the neglect, including normal power. He was meningitic, but not distressed, and afebrile. No ictal activity was observed. There was residual right parotid swelling. Ocular, cardiac, respiratory, abdominal, skin and joint examinations were normal.

HIV antibody was positive, with a viral load of 500 000 copies/ml and the CD4 count was 94 cells/mm3. ESR was 120 mm/h. Serum amylase was mildly elevated at 215 U/l; a parotid origin of this could not be confirmed. A homogenous staining antinuclear antibody was weakly positive at 1:100, increasing after 1 week to >1:1000. A weakly staining antineutrophil cytoplasmic antibodies was positive, but testing for myeloperoxidase and PR3 antigens were negative, along with other autoimmune markers. Serial blood cultures were negative. CSF showed 18 nucleated cells, mainly lymphocytes with some macrophages, consistent with a reactive pleocytosis. CSF protein was elevated at 0.72 g/l and glucose 3.4 mmol/l (serum 4.4 mmol/l). Chest radiography and abdominal ultrasonography were normal. Transthoracic and transoesophageal echocardiography showed normal cardiac valves.

MRI brain with gadolinium enhancement demonstrated numerous lesions in both cerebral hemispheres with a preponderance of the left hemisphere (fig 1A–D). The large left occipital lesion had acutely restricted diffusion on diffusion weighted imaging and ADC images (not shown). The clinical impression was of a cerebral vasculitis in the context of HIV infection. Brain biopsy was performed of the superficial left occipital lesion to exclude additional pathology such as concomitant fungal infection. The biopsy showed only non-specific necrosis, with no specific features of a vasculitis. There was also no evidence of emboli, lymphoma, fungi, viral inclusions or mycobacteria identified. Cerebral angiography (fig 1E) confirmed a macro and microangiopathic vasculopathy. PCR of both CSF and biopsied brain tissue was negative for varicella zoster virus, herpes simplex virus, JC virus, Epstein–Barr virus, cytomegalovirus, human herpesvirus six, adenoviridae and enteroviridae. Mycobacterial investigations were negative (cultures of blood and CSF along with CSF Tb PCR). Testing of blood and CSF was also negative for Aspergillus, Cryptococcus, Toxoplasma, Treponema, Borrelia and current hepatitis A, B or C.

Figure 1

(A, B) FLAIR MRI axial images showing multiple hyperintensities with preponderance for the left cerebral hemisphere. (C, D) T1 gadolinium enhanced axial MRI images showing additional enhancing lesions not seen on the T2 and FLAIR sequences. (E) Digital subtraction angiography after left internal carotid artery injection. There is delayed filling of the left middle cerebral artery (MCA) territory, shown in the early arterial phase having mostly filled via leptomeningeal collaterals from the left anterior cerebral artery (ACA). The ACA territory is entering the venous phase, having completed arterial filling earlier in the sequence. This suggests stenosis of proximal MCA segments. In addition, there is evidence of microangiopathy with segmental stenosis of MCA arterioles indicated by the arrows.

Empiric intravenous acyclovir was started on presentation but discontinued when viral PCR testing returned negative and there had been no clinical improvement. HAART was initiated (Fosamprenavir, Ritonavir, Kivexa) with cotrimoxazole for pneumocystis prophylaxis. He made a slow clinical improvement, walking at 3 weeks and oriented at 6 weeks, and was able to be discharged home. He has subsequently returned to work. MRI at 2 months showed maturation of the larger lesions, with no new abnormalities. We did not find earlier reports of major clinical improvement in HIV associated cerebral vasculopathy or vasculitis with HAART alone, without the use of immunosuppressive treatment.

The distribution of the cerebral lesions on MRI, the macro and microangiopathy on cerebral angiography, the CSF pleocytosis and the reversible encephalopathy together suggest that the vasculopathy here is most likely due to a cerebral vasculitis. The brain biopsy contained necrotic material consistent with an infarct and therefore did not prove a vasculitis. The elevated antinuclear antibody was in isolation, with no other clinical or serological features of a primary rheumatological disorder. HIV was the only cause identified, and there was a clinical response to treatment for this.

HIV associated intracranial vasculopathy is recognised in association with young onset stroke, found in six of 67 young onset stroke patients in a recent series.2 Biopsy proven HIV associated cerebral vasculitis has been identified after initiation of HAART, and is thought in these cases, at least to some extent, to be part of an immune reconstitution inflammatory syndrome.3 4 This often requires temporarily stopping the antiretroviral treatment and additional immunosuppression. In the case we describe, the neurological presentation was prior to the diagnosis of HIV and the initiation of antiretroviral treatment.

It can be tempting in such cases to use immunosuppressive treatments, as in a primary CNS vasculitis, but this has theoretical risks of exposing the already immunocompromised patient to additional risk. We propose that thorough investigation for other causes, along with careful clinical observation after initiation of HAART, was the appropriate management in this case with an excellent recovery.



  • Competing interests: None.

  • Patient consent: Obtained.

  • NJC, HS and MWW are formerly of the Neurology Department, Royal Free Hospital, London, UK, where the patient described here was treated.

  • See Editorial Commentary, p 831

Linked Articles