The title of this poster presentation does not accurately reflect the design and conduct of this trial. This study would be better described as a partial treatment interruption trial, and an unorthodox one at that. But in defense of the designers, this must have been planned when hydroxyurea (HU, Hydrea) was in fashion for the treatment of HIV therapy.

One hundred and eighty-seven HIV-infected patients were studied in this protocol. There were 69 patients on protease inhibitor (PI)-based therapy and 116 on non-nucleoside reverse transcriptase inhibitor (NNRTI)-based therapy. These patients had to have CD4 counts over 350 cells and HIV-RNA (viral load) lower than 50 copies/mL in order to be eligible for the study. They did not have a "drug holiday," but rather had their treatment switched to a combination of ddI (didanosine, Videx) 400 mg daily and HU 500 mg BID.

After this therapy switch, 21 percent of the patients experienced a viral load rebound above 10,000 copies/mL (19 on PIs, 20 on NNRTIs). These patients with rebound were then restarted on a HAART regimen. Among these patients with rebound, those who restarted a PI-based regimen all had suppression to less than 50 copies/mL. However, 16/29 patients who restarted an NNRTI-based regimen did not attain an HIV-RNA level of less than 50 copies/mL.

When the resistance data from the NNRTI-treated patients was examined, it provided a partial explanation for this high rate of failure. As expected, some of the patients who had viral rebound on ddI/HU had evidence of multiple NRTI mutations, and these patients were much less likely to respond to re-initiation of an NNRTI-based therapy. In essence, many of these patients received effective monotherapy with an NNRTI.

The group of 13 patients who did re-suppress with an NNRTI-based regimen had wild type virus, which explains their favorable response to therapy.

This study really does not add much to our current understanding of structured treatment interruption (STI). It was designed a while back before the considerable toxicity with HU was fully recognized. Moreover, few researchers would be eager now to embark on a study that had patients on therapy with well-controlled viral replication switch to a non-suppressive regimen. The risk of development of viral resistance would be too high, as this study aptly demonstrated.