Is the diagnosis at hand?

Abstract

Listen

Translate article

In August, 2008, a 49-year-old west African woman presented to us with a 1-month history of blurred vision, dizziness, and falls. A friend had noticed increasing confusion and unusual behaviour, and found her collapsed at home. She was diagnosed with HIV-1 infection in August, 2007, when the patient presented with fungal nail infection. At diagnosis, HIV VL was 1 967 745 copies per mL with a CD4-cell count of 22 per μL (3·8%) and Pneumocystis jirovecii prophylaxis and combination antiretroviral therapy (cART) were started. Within 12 weeks, VL became undetectable (<40 copies per mL) with a CD4-cell count of 585 per μL (14·6%). During initial clinic attendances, she was quiet and slightly withdrawn, which we attributed to her personality. In April, 2008, she developed HIV viraemia attributed to poor cART concordance. Genotypically confi rmed drug resistance prompted cART switch. On examination, she was apyrexial and normotensive. She was disoriented in time, person, and place. She showed subcortical slowing, perseveration, and marked working memory impairment without focal neurological defi cit. Fundoscopy was normal. Investigations showed: HIV VL 12 062 copies per mL; CD4-cell count 502 per μL (16·1%); normal C-reactive protein concentration, white cell count, and lactate dehydrogenase. Treponema serology and cryptococcal antigen were negative. Brain MRI (fi gure A) showed extensive, multifocal signal abnormalities. Crowding of the cerebellar tonsils at the foramen magnum precluded lumbar puncture. The diff erential diagnoses were: viral encephalitis; acute disseminated encephalomyelitis; diff usely infi ltrative neoplasia; or immune reconstitution infl ammatory syndrome (IRIS) of pre-existing progressive multifocal leucoencephalopathy, toxo plasmosis, or HIV-associated neurocognitive disorder (HAND). She was treated empirically for herpes encephalitis and cerebral toxoplasmosis; montelukast was administered for possible IRIS. cART was continued. To avoid masking of lymphoma, steroids were withheld. On day 10, she was less disoriented, but repeat MRI showed progression of disease. A brain biopsy sample taken on day 24 showed perivascular large CD20-positive B cells (fi gure B) with 25% proliferation fraction, but Epstein-Barr virus (EBV) encoded RNA negative on in-situ hybridisation. There was also marked peri vascular and parenchymal T-cell infi ltration (pre dominantly CD8-positive) and microgliosis (CD68-positive). Tissue immunocytochemistry was negative for all viruses including EBV and HIVp24. She con tinued to improve without steroids, and repeat MRI (day 34) showed improvement (fi gure C). Given the clinical evolution and histopathology, a diagnosis of an IRIS driving HAND was made instead of B-cell lymphoma. She was discharged attaining full independence. When seen in November, 2009, she was fully interactive and spontaneously initiated appro priate conversation. HAND is a prevalent cause of morbidity in patients with HIV infection. Patients present with subcortical impairment manifested by slowed motor speed and decreased atten tion. Diagnosis is by exclusion and risk factors include low nadir CD4-cell count, apolipoprotein E4 carriage, and age. IRIS occurs in about a third of patients after cART initiation and probably represents immune hyperactivation towards foreign or self antigen. Diagnostic criteria include: worsening infl ammatory symptoms unexplained by new or previous infection, disease, or drug toxicity; temporal relation with cART; and ≥1 log fall in HIV VL. IRIS is asso ciated with severe immunosuppres sion. Histopatho logical studies of brain-IRIS emphasise CD8-positive T-cell lympho cytosis (CD4expressing T cells absent/rare) and CD68-positive microgliosis. Despite limited evidence, cortico steroids, non-steroidal anti-infl ammatory drugs, and monte lukast have been tried. In retrospect, our patient prob ably had features of HAND at HIV diagnosis. cART triggered IRIS, un masking HAND and precipitating cART non-con cordance. We propose that patients presenting with ad vanced HIV infection at diagnosis have formal neuro cognitive testing (±brain MRI) to assess for subclinical HAND.