Changing disease patterns in focal brain lesion-causing disorders in AIDS.

Objectives: To assess temporal trends of the different disorders causing focal brain lesions (FBL) in HIV-infected patients and to examine the reliability of the U.S. Centers for Disease Control and Prevention (CDC) criteria for presumptive diagnosis of toxoplasmic encephalitis (TE) for the years 1991 to 1996.
Design/methods: A prospective, monocenter study. Percentages of occurrence of the different FBL-causing disorders for each year were calculated. Temporal trends were analyzed by chi2 test for linear trend and multivariate polytomous nonordinal logistic regression. The positive predictive value (PPV) of the CDC's presumptive criteria for the diagnosis of TE (recent onset of a focal neurologic abnormality consistent in intracranial disease or a reduced level of consciousness, evidence on brain imaging of a lesion having mass effect or the radiographic appearance of which is enhanced by injection of contrast medium, and serum antibody to toxoplasmosis) was calculated using contingency tables for each calendar year.
Results: A highly significant decline of the risk of TE and an increase of the probability of patients to take anti-Toxoplasma prophylaxis were observed. A threefold but statistically not significant augmented risk of diagnosing both primary central nervous system lymphoma (PCNSL) and progressive multifocal leucoencephalopathy (PML) has been registered for 1996 compared with 1991. Among FBL showing contrast enhancement, the increased finding of PCNSL over the years studied was significant. The probability of other FBL-causing disorders, such as focal viral encephalitis sustained by cytomegalovirus or herpes simplex virus, increased significantly over the years studied. Multivariate analysis confirmed that the year of diagnosis of FBL had a significant effect on the risk reduction of TE. The PPV of the CDC's criteria for the presumptive diagnosis of TE dropped from 100% for the year 1991 to 39% in the year 1996. A similar result was obtained in calculating the PPV of presumptive criteria only among patients without previous primary prophylaxis.
Conclusions: Because of the significant decrease of TE and the increase of PCNSL empiric anti-Toxoplasma therapy no longer seems appropriate as a first-line approach to all HIV-positive patients with FBL. Especially in the case of a finding of FBL by contrast enhancement, new diagnostic strategies should be employed to identify the underlying disorder rapidly and accurately.