Recent studies have shown that patients receiving state-of-the-art treatment for HIV infection (highly active antiretroviral therapy, or HAART) show discordant responses to therapy-that is, they improve both clinically and immunologically (increased CD4 immune cell counts), but their levels of HIV in the blood do not drop. This study will examine blood samples taken from HIV-infected children treated with 2',3'-dideoxyinosine (didanosine, or ddI) in an earlier NCI study to determine if these patients had similar discordant responses to therapy. If so, the study will also examine how these factors may be predictive of disease progression and survival.

Investigators will measure HIV levels in blood specimens from children previously enrolled in NCI protocol 88-C-0129. The measurements will be done in specimens drawn before initiation of treatment with ddI and at various times during the course of treatment. The viral responses to therapy will then be correlated with short- and long-term clinical and immunologic outcomes.

In addition, researchers will examine these patients' blood samples for additional factors recently discovered to potentially influence disease progression. They include selenium levels and certain genetic factors.

A better understanding of discordant responses to therapy will help physicians determine the best treatment option in situations where therapy results do not show uniform benefit.

This protocol involves the scientific examination of laboratory specimens only and is not a clinical study open to patient enrollment.

This study will examine clinical and immunologic data and HIV-1 RNA levels in banked serum specimens from HIV-infected pediatric patients previously enrolled on NCI protocol 88-C-0219, a phase I/II study of 2', 3'-dideoxyinosine (ddI). In the original study serum p24 antigen was used as the direct measurement of viral quantitation. Advances in understanding of HIV pathophysiology have demonstrated that HIV-1 RNA level as measured by quantitative RNA PCR is a far superior molecular tool for assessing viral replication and burden in vivo. Analysis of clinical, HIV-1 RNA and CD4 responses will be performed to identify patients with discordant surrogate marker responses to therapy, defined as: improvement in immunologic parameters in the setting of virologic non-response or failure. Correlation will then be made with short and long term clinical outcomes.

Following the initial analysis to identify patients with discordant responses, selenium levels, viral genotypic and phenotypic profiles of resistance to ddI, and in vitro viral fitness will be examined. These biologic parameters have recently been determined to potentially play an important role in clinical, immunologic and virologic HIV-related outcomes. Parameters will be examined in subsets of patients with the following surrogate marker response profiles: a.) immunologic and virologic responders; b.) immunologic and virologic non-responders; and c.) discordant immunologic and virologic responders. Specimens from the ddI monotherapy study were chosen for this analysis because a substantial body of clinical and immunologic data is available on patients up to 7 years.

INCLUSION CRITERIA:

Prior enrollment on NCI protocol 88-C-0219, a phase I/II study of 2',3'- dideoxyinosine (ddI) in children with human immunodeficiency virus infection.

Informed consent for participation in 88-C-0219 clinical trial.

Adequate available clinical and immunologic data and banked blood specimen components obtained during implementation of the protocol.