To compare the safety and effectiveness of orally administered didanosine (ddI) with high dose orally administered zidovudine (AZT) in patients who develop or exhibit progression of the AIDS dementia complex (ADC) and who have not previously been intolerant to AZT at doses of up to 1000 mg/day.

HIV-infected or AIDS patients may develop ADC which causes damage to the nervous system. ADC may be caused by some action of the AIDS virus on the nervous system, although similar problems can be caused by other infections because the AIDS virus lowers the body's ability to fight other infections. It is important to determine whether symptoms are due to ADC or to some other infection since treatment varies for different conditions. AZT has been shown to be beneficial to people with ADC although its effectiveness has only been studied in a small number of patients. Studies suggest that higher doses of AZT are more likely to be effective than standard doses in improving symptoms of ADC.

HIV-infected or AIDS patients may develop ADC which causes damage to the nervous system. ADC may be caused by some action of the AIDS virus on the nervous system, although similar problems can be caused by other infections because the AIDS virus lowers the body's ability to fight other infections. It is important to determine whether symptoms are due to ADC or to some other infection since treatment varies for different conditions. AZT has been shown to be beneficial to people with ADC although its effectiveness has only been studied in a small number of patients. Studies suggest that higher doses of AZT are more likely to be effective than standard doses in improving symptoms of ADC.

Patients are randomly assigned to receive either oral ddI or oral AZT.

Inclusion Criteria

Concurrent Medication:

Allowed:

• Chronic suppressive therapy for herpes simplex virus, cytomegalovirus, Candida albicans, and Salmonella.
• Prophylactic therapy for Pneumocystis carinii pneumonia.
• Maintenance anticonvulsant therapy following a seizure in the context of the AIDS dementia complex.
• Isoniazid only if no acceptable alternative therapy is available.
• Judicious use of benzodiazepines, tricyclics, and other antidepressants is allowed but a stable dose level should be obtained prior to entry and maintained throughout the trial.
• In patients for whom it is medically necessary to initiate or alter therapy with these drugs during the initial 16 week study period, data will not be used in the study.
• Metronidazole for single courses of therapy not to exceed 14 days within consecutive 90-day intervals, the first of which begins at the initiation of the study.
• Erythropoietin for patients under the relevant Treatment IND.
• Symptomatic therapies (such as analgesics, antihistamines, antiemetics, antidiarrheal agents).

Allowed but not encouraged:

• trimethoprim /sulfamethoxazole (T/S) or other sulfonamides.

Patients must have the following:

• Screened for other causes of dementia.
• Stage 1, 2, or 3 AIDS dementia complex.
• Estimated premorbid IQ of at least 70.
• Anti-HIV antibody or HIV in blood and/or cerebrospinal fluid.
• If prior history of positive syphilis serology, should have been treated with appropriate course of antibiotics; if not, such treatment should be administered prior to pretreatment screening.
• Not have previously shown intolerance to zidovudine (AZT).
• Able (or parent and/or guardian able) to provide written consent.

Allowed:

• Basal cell carcinoma, in situ carcinoma of the cervix, Kaposi's sarcoma without evidence of visceral involvement or not requiring systemic chemotherapy.

Exclusion Criteria

Co-existing Condition:

Patients with the following conditions or symptoms are excluded:

• Grade 3 neuropathy, based on the Neuropathy Targeted Symptom.
• Questionnaire, or patients with any moderate abnormality indicative of peripheral neuropathy including stocking loss of sensation (to sharp pain, light touch, or vibration), distal extremity weakness (< 4/5), or absent ankle jerks.
• History of present or past acute or chronic pancreatitis.
• Active, symptomatic AIDS-defining opportunistic infection and requiring any ongoing maintenance therapy for confounding neurologic disease.
• Severe premorbid psychiatric illness including bipolar illness, schizophrenia, and electroconvulsive therapy.

Previous neurological disease unrelated to HIV infection:

• multiple sclerosis, documented stroke, degenerative disease.
• Patients with chronic seizure disorders or head injury will only be excluded if the condition results in functional impairment or is likely to interfere with the evaluation.
• Concurrent or previous central nervous system infections or neoplasms as revealed by Computerized Tomography (CT) or Magnetic Resonance Imaging (MRI) scan or cerebrospinal fluid analysis (such as toxoplasmosis, primary or metastatic Central Nervous System (CNS) lymphoma, progressive multifocal leukoencephalopathy, cryptococcal or other fungal meningitis, tuberculous Central Nervous System (CNS) infections, and untreated neurosyphilis).

Concurrent Medication:

Excluded:

• Intravenous pentamidine. DHPG (Ganciclovir) should not be co-administered.
• Monoamine oxidase (MAO) inhibitors, phenothiazines, butyrophenones, barbiturates, amphetamines.
• Oral acidifying agents.

Patients with the following are excluded:

• Neoplasms not specifically allowed.
• Grade 3 neuropathy.
• History of present or past acute or chronic pancreatitis.
• Active, symptomatic AIDS-defining opportunistic infection.
• Requiring any ongoing maintenance therapy for confounding neurologic disease.
• Conditions listed under Exclusion Co-existing Conditions.

Prior Medication:

Excluded within 30 days of study entry:

• Anti-HIV therapy other than zidovudine (AZT).
• Biologic response modifiers.
• Corticosteroids.
• Drugs toxic to peripheral nerves.
• Investigative drugs.
• Neurotoxic drugs.

Excluded:

• Dideoxycytidine (ddC).

Active alcohol or drug abuse or methadone maintenance sufficient, in the investigator's opinion, to prevent adequate compliance with study therapy and evaluations.