To determine whether cerebrospinal fluid (CSF) viral burden measurements can assist in the evaluation of human immunodeficiency virus (HIV)-associated neurocognitive disorders, we quantified HIV type 1 (HIV-1) RNA in CSF. Because previous findings suggested that disease stage, lymphocytic pleocytosis, and HIV-1 RNA levels in plasma may influence CSF viral burden, these variables were examined as potential modifying factors. HIV-1 RNA levels were quantified by using a reverse transcriptase-polymerase chain reaction assay. Performance on a comprehensive neuropsychological (NP) battery was noted in 97 prospectively enrolled, HIV-infected subjects. Among subjects with acquired immunodeficiency syndrome (AIDS) (<200 CD4+ lymphocytes), NP impairment was associated with significantly higher CSF RNA levels (3.1 vs 1.8 log10 copies/ml; p = 0.02); most impaired subjects met criteria for HIV-associated dementia or minor cognitive-motor disorder. In subjects without AIDS, CSF RNA and NP impairment were unrelated. Before AIDS, CSF RNA was strongly correlated to plasma RNA and to pleocytosis, but in AIDS, CSF and plasma RNA were independent. In conclusion, we found elevated CSF HIV-1 RNA levels in NP impaired subjects with AIDS. Before AIDS, systemic viral replication, possibly through CD4+ mononuclear cell trafficking, may govern virus levels in CSF, whereas in AIDS, CD4 cell depletion may unmask a correlation between increased productive central nervous system HIV infection and clinical neurocognitive disorders.

The HIV-infected individuals with NP impairment had a higher risk of dying than those without impairment. This was particularly true for those meeting syndromic diagnostic criteria.

There is a high prevalence and diversity of non-B HIV subtypes in this large cohort. Efficient diagnosis of acute primary HIV-1 infection was identified as a goal for prevention programs. Modifiable risk behaviors and target populations for intervention were identified.

We studied the pedigrees of 17 index patients with osteosarcoma, recording malignant disease and cause of death for first- and second-degree relatives. There were seven cancers and five cancer deaths per 2151.5 person-years in first-degree relatives of osteosarcoma patients under the age of 50 years, a significantly greater incidence than in an age- and sex-matched population group (p < 0.001). This excess of malignancy was largely due to two families which fulfilled the criteria for the Li-Fraumeni cancer family syndrome. Both of these families were shown to have the genetic alterations in the p53 gene which have been implicated in this syndrome. Our study suggests that orthopaedic surgeons seeing new cases of osteosarcoma should arrange screening for familial malignancy.

Sexually transmitted diseases in HIV-infected subjects are correlated with neurocognitive impairment through an unidentified mechanism.