Intracerebral Lesions in HIV Patients

ABSTRACT & COMMENTARY

Source: Sadler M, et al. Management of intracerebral lesions in patients with HIV: A retrospective study with discussion of diagnostic problems. Q J Med 1998;91:205-217.

About half of all aids patients sooner or later develop clinical neurologic complications; almost all patients show neurologic abnormalities on autopsy. Sadler et al review the diagnosis and management of a group of 95 patients of a London HIV center who suffered from various CNS diseases related to HIV infection. (Total annual patient attendance at the Center equaled approximately 2100.) Of the 95, 73 had proved or presumed specific CNS diseases. The largest group had firm or presumed evidence of toxoplasma encephalitis (TE; n = 32). The other two diagnosable abnormalities consisted of primary CNS lymphoma (PCNSL; n = 17), or progressive multifocal encephalitis (PML; n = 24). Twenty patients had neurologic symptoms but could not be specifically diagnosed as to the cause. Two had unrelated disease. All patients received initial brain imaging by CT or MRI. Additional diagnostic procedures included annual measurement for toxoplasma antibodies as well as polymerase chain reaction (PCR) testing for viruses in spinal fluid.

Brain images in patients with TE typically included more than five lesions with ring enhancement surrounded by edema and mass effect affecting basal ganglia and cortical gray-white matter borders. PCNSL, by comparison, produced fewer lesions, a more homogenous contrast enhancement, adjacency to ependymal surfaces, and extension across the corpus callosum. PML generated high signal on T-2 weighted images with hypodensity of the same areas on T-1 images.

TE was diagnosed as "definite" or "presumed" by: 1) the appearance of specific or non-specific changes in brain imaging; 2) clinical findings that included incoordination and other motor defects reflecting focal cerebral involvement, headache (29%), and seizures; and 3) TE serum antibody titers greater than 1:16. Improvement of brain scans and clinical signs following two weeks of treatment strongly implied intracranial toxoplasmosis, whereas continued neurological deterioration suggested the presence of PCNSL. Toxoplasma treatment consisted of sulphadiazine and pyrimethamine (first choice), with clindamycin substitutes if sulphadiazine was not tolerated. Atovaquone replaced the above drugs under condition of complete intolerance. Most patients with TE improved within up to two weeks, with occasional delays up to six weeks.

Clinical signs of PCNSL generally resembled those of TE. Definitive diagnosis of PCNSL rested on either finding a PCR of spinal fluid identifying Epstein-Barr virus or the product of a brain biopsy. Thallium brain scanning can also identify the tumor cells that represent PCNSL. PML lesions produce high white matter enhancement on MR images and are usually associated with focal motor abnormalities, visual deficits, confusion, and, sometimes, seizures. Explicit diagnosis can be attained by PCR identification of Creutzfeld-Jakob (CJ) virus in CSF in 90-100% of patients.

Ultimately, the most significant item in this report is the importance of early treating cases of intracranial TE. Reflecting other similar reports, Sadler et al plotted median post-treatment survival times for definitive TE at 528 days (34-746+) and for presumed TE at 33 (0-67) days.

The authors emphasize that, while distinction between TE and PCNSL may be difficult acutely, TE is a far more treatable disease than PCNSL. Accordingly, in the presence of cerebral mass lesions of uncertain cause, patients possibly having TE should be started on a trial of anti-toxoplasma as early as possible. Particularly favored for such early therapy of TE are serum antibodies of greater than 1:16 (several studies indicate that TE is unlikely in patients with intracranial lesions and lower TE titers). Since such antibodies can also be present in patients with PCNSL, further deterioration or failure to improve clinically or by brain imaging should accentuate efforts to identify the PCNSL by PCR analysis of CSF for Epstein-Barr virus or by direct brain biopsy.

By comparison, mean and extreme survival days after definitive (D) or presumptive (P) diagnosis were, for PCNSL, D = 24 (11-37) and P = 73 (27-119). For patients definitely diagnosed for PML, survival lasted a mean of 30 days (range, 22-38), whereas patients presumptively diagnosed survived 75 days (range, 41-109). These latter two groups are small, but their outcomes resemble those in other series.

COMMENTARY

This report precedes any changes in diagnosis or treatment of toxoplasma encephalitis, primary CNS lymphoma, and multifocal leucoencephalopathy that may come from the wide application of new protease and other antiretroviral drugs in AIDS. Clear evidence from several more up-to-date sources indicates that cases of AIDS-dementia in the United States during the past 2.5 years have declined precipitously, as has the admission-to-hospital rate for AIDS in general. A consortium of French centers showed a lessening of hospital days by 35%, a drop of new cases by 35%, and a drop of deaths by 46% (Moutau Y, et al. AIDS 1997;11:101-105). Specific antiretroviral consumption rose by 53%.

The International AIDS Society-USA Panel, faced with the recently obtained remarkable remissions of AIDS symptoms in patients taking antiretroviral therapy, has recently emphasized the need for earlier initiation and more aggressive therapy for AIDS than that recommended a year ago (Carpenter CCJ, et al. JAMA 1997;277:1962-1969). For those following such patients, the tested viral load is crucial. Careful selection of new therapeutic drugs plus frequent appraisals of potentially rising viral load guides whether new, more aggressive drugs are desirable. -fp