To compare the proportion of patients in the 2 zidovudine (ZDV)-containing arms who have a plasma HIV RNA concentration below the limit of detection (defined as 500 copies/ml or less) at Weeks 20 and 24 [AS PER AMENDMENT 8/24/98: HIV RNA concentration below the limit of detection is now defined as 200 copies/ml or less]. To compare the safety and tolerability of the different treatment regimens. To compare the decrease in plasma HIV-1 RNA and the change in CD4 count from baseline to the average of Weeks 20 and 24 [AS PER AMENDMENT 12/19/97: and to the average of Weeks 44 and 48; AS PER AMENDMENT 8/24/98: and the average of Weeks 88 and 96] in the 2 ZDV-containing arms. To study the emergence of resistance to ZDV, lamivudine (3TC), stavudine (d4T), delavirdine (DLV), and indinavir (IDV) in treated patients. To correlate the antiviral and immunologic activity and emergence of drug resistance with pharmacologic parameters of study drugs. To delineate the pharmacokinetic interactions of IDV and DLV. [AS PER AMENDMENT 12/19/97: To delineate the possible development of cellular resistance to nucleoside analogs and the consequences of switching nucleoside study drugs on intracellular phosphorylation.] To document rates and patterns of adherence over the course of the study, from day of randomization through 48 weeks. [AS PER AMENDMENT 8/24/98: To define long-term durability of the virologic activity of the different treatment regimens, as defined by the proportion of patients with plasma HIV-1 RNA levels that remains below the limit of detection. To define long-term tolerability of the different treatment regimens.] Although a change in reverse transcriptase (RT) inhibitors is recommended when adding or changing protease inhibitors in a treatment regimen, the choice of available RT inhibitors is often limited by prior exposure, toxicity, or pharmacologic interaction with the protease inhibitors. This study addresses the question of whether to continue 3TC or substitute the nonnucleoside reverse transcriptase inhibitor (NNRTI) DLV when adding IDV to therapy for patients previously treated with ddI or d4T plus 3TC who have greater than 500 copies/ml of plasma HIV-1 RNA. Although the activity of DLV as monotherapy or in combination with nucleoside reverse transcriptase inhibitors is of limited duration due to rapid emergence of resistance, it is possible that DLV will contribute significantly to the activity of 3-drug regimens that include a new RT inhibitor plus a protease inhibitor.

Although a change in reverse transcriptase (RT) inhibitors is recommended when adding or changing protease inhibitors in a treatment regimen, the choice of available RT inhibitors is often limited by prior exposure, toxicity, or pharmacologic interaction with the protease inhibitors. This study addresses the question of whether to continue 3TC or substitute the nonnucleoside reverse transcriptase inhibitor (NNRTI) DLV when adding IDV to therapy for patients previously treated with ddI or d4T plus 3TC who have greater than 500 copies/ml of plasma HIV-1 RNA. Although the activity of DLV as monotherapy or in combination with nucleoside reverse transcriptase inhibitors is of limited duration due to rapid emergence of resistance, it is possible that DLV will contribute significantly to the activity of 3-drug regimens that include a new RT inhibitor plus a protease inhibitor.

Patients with greater than 500 HIV-1 RNA copies/ml are randomized to 3 treatment arms as follows:

Arm I: d4T + ZDV placebo + DLV + IDV Arm II: ZDV + d4T placebo + 3TC + IDV Arm III: ZDV + d4T placebo + DLV + IDV Treatment on all arms is given for 24 weeks. [AS PER AMENDMENT 12/19/97: The study is no longer partially blinded, and placebo agents are no longer given; treatment duration is now 48 weeks.] [AS PER AMENDMENT 8/24/98: study duration is now 96 weeks.] Rollover patients from ACTG 306 with greater than 500 HIV-1 RNA copies/ml previously assigned to ZDV/3TC are nonrandomly assigned to Arm I; those previously assigned to ddI/3TC or d4T/3TC are randomized to Arm II or III. Non-rollover patients are randomized to Arm II or III. Rollover patients from ACTG 306 with 500 HIV-1copies/ml or less continue on their previously assigned regimen [AS PER AMENDMENT 12/19/98: current regimen must be ZDV/3TC, ddI/3TC, or d4T/3TC.] for the study duration or until an increase occurs. If this increase occurs, patients previously assigned to ZDV/3TC are nonrandomly assigned to Arm I for the remaining study weeks, while those previously assigned to either ddI/3TC or d4T/3TC are randomized to Arm II or III for the remaining study weeks. Patients who received ddI/d4T or ddI/3TC in ACTG 306 are stratified by whether patients received monotherapy or combination therapy during the first 24 weeks [AS PER AMENDMENT 12/19/97: 48 weeks]; [ AS PER AMENDMENT 8/24/98: 96 weeks.] of ACTG 306.

Inclusion Criteria

Concurrent Medication:

Required:

• Patients completing ACTG 306 who remain on blinded therapy through the extension period or
• Patients on stable (6 months or greater) ddI/3TC or d4T/3TC combination therapy who have plasma HIV-1 levels higher than 500 copies/ml by the Amplicor HIV-1 Monitor Assay.

Allowed following contact with Protocol Pharmacologist:

• Diltiazem, nifedipine, phenytoin, and warfarin.

Patients must have:

• Absolute CD4 count of 200 cells/mm3 or greater.
• HIV-1 RNA levels greater than 500 copies/ml by the Amplicor HIV-1 Monitor assay. NOTE:
• This is a requirement for those receiving study medication. [AS PER AMENDMENT 12/19/97:
• HIV-1 infection must be documented by any licensed ELISA test kit and confirmed by either Western blot, HIV culture, HIV antigen, plasma HIV RNA, or a second antibody test by a method other than ELISA at any time prior to entry.]
• Signed, informed consent from a parent or legal guardian for patients under 18 years of age.
• Life expectancy of at least 24 weeks.

Exclusion Criteria

Co-existing Condition:

Patients with the following symptoms or conditions are excluded:

• Unexplained temperature of 38.5 C or higher for 7 consecutive days, or chronic diarrhea defined as more than 3 liquid stools per day persisting for 15 days, within 30 days prior to study entry.
• Proven or suspected acute hepatitis within 30 days prior to study entry.
• Malignancy that requires systemic chemotherapy. NOTE:Patients with minimal Kaposi's sarcoma (KS) fewer than 5 cutaneous lesions and no visceral disease or tumor-associated edema) are allowed to enroll provided that they do not require systemic therapy.

Concurrent Medication:

Excluded:

• Concurrent ZDV (for patients other than those rolling over from ACTG 306).
• Any experimental antiretroviral agents or other experimental therapies.
• Acute therapy for an infection or other medical illnesses within 14 days prior to study entry.
• Recombinant erythropoietin (rEPO), G-CSF, or GM-CSF within 30 days prior to study entry.
• Interferons, interleukins, or HIV vaccines within 30 days prior to study entry.
• Rifampin, rifabutin, cisapride, triazolam, midazolam, terfenadine, astemizole, or loratadine, within 14 days prior to study entry.

Patients with the following prior conditions are excluded:

• History of acute or chronic pancreatitis.
• History of Grade 2 or higher bilateral peripheral neuropathy. [AS PER AMENDMENT 12/19/97: Patients with Grade 2 or 3 peripheral neuropathy due to current use of ddI/3TC or d4T/3TC and who have a screening viral load above 500 copies/ml are eligible as they will be randomized to a regimen that does not contain an agent associated with peripheral neuropathy toxicity.]

Prior Medication:

Excluded:

• Prior NNRTI or protease inhibitor therapy.
• Prior ZDV (for patients other than those rolling over from ACTG 306).
• Previous induction or maintenance therapy with foscarnet.