The purpose of this study is to compare 2 treatment plans to try to increase the effects of anti-HIV drugs in patients who are resistant to the drug effects.

Sometimes the increase in a patient's viral load (the level of HIV in the blood) can be slowed or stopped by taking anti-HIV drugs. This does not always happen. Sometimes anti-HIV drugs work at first but then stop working. When most of the usual anti-HIV drugs no longer seem to work, the virus is called multidrug-resistant (MDR). This study will compare 2 treatment plans to try to increase the effects of anti-HIV drugs in patients with MDR virus.

An increasing number of patients are developing multidrug-resistant (MDR) virus, as determined by genotypic antiretroviral resistance testing (GART), due to treatment failure to suppress viral replication after several rounds of combination antiretroviral therapy. The best therapeutic strategy for these patients is uncertain. Two strategies currently being used are (1) STI followed by a new antiretroviral regimen and (2) immediate initiation of a new antiretroviral regimen.

Patients are screened for the presence of MDR virus and a plasma HIV RNA level greater than 10,000 [AS PER AMENDMENT 07/03/01: greater than 5,000] copies/ml. Eligible patients attend a baseline visit [AS PER AMENDMENT 07/03/01: and a subsequent randomization visit] where the qualifying GART results are provided. Patients who consent to participate have phenotypic antiretroviral resistance testing (PART) done on a specimen from the same blood draw that was used for the GART evaluation. After PART results are available, patients are randomized [AS PER AMENDMENT 07/03/01: If the predicted sensitivities are not available for some or all drugs included in the PART, the patient may still be randomized.] to either a 4-month STI followed by a new antiretroviral regimen or an immediate new antiretroviral regimen. The antiretroviral regimens chosen are based on the patients' history and both GART and PART results. [AS PER AMENDMENT 07/03/01: Additional GART and PART may be requested after at least 4 months of antiretroviral treatment.] Patients have the follow-up data collection done at Months 1-8 and every 4 months thereafter. Changes in antiretroviral therapy, Grade 4 adverse experiences, progression of disease, and deaths are reported as they occur. Patients are seen for clinical management as often as deemed necessary. All patients are followed to a common closing date estimated to be 24 months after the last patient is randomized. Some patients may participate in a Point Mutation Substudy [AS PER AMENDMENT 07/03/01: Plasma Point Mutation Substudy and PBMC Point Mutation Substudy].

Inclusion Criteria

Patients may be eligible if they:

• Have proof of MDR virus from a blood test.
• Have a viral load above 5,000 copies/ml from the same blood sample showing MDR virus.
• Intend to start a new anti-HIV treatment around the time of the study.
• Have been on a stable anti-HIV treatment between 14 days prior to the blood test mentioned above and when they are randomly assigned to a treatment.
• Are at least 13 years old (consent of parent or guardian required if under 18).
• (This protocol has been changed to reflect new criteria.)

Exclusion Criteria

Patients will not be eligible if they:

• Have received a vaccine or had an illness that might affect viral load within 14 days before the blood test showing MDR virus.
• Have received IL-2 within 4 months of the above-mentioned blood test or plan to take IL-2 during the study.
• Have an opportunistic (AIDS-related) infection requiring treatment.
• Are pregnant or breast-feeding.
• Are currently participating in CPCRA 057 (PIP study).