The impetus to explore the potential utility of intermittent HIV therapy stems from observations made by Dr. Franco Lori and colleagues involving a person now commonly referred to as the "Berlin patient". This 29-year-old man began treatment with ddI (Videx®), hydroxyurea, and Crixivan® within about a month of presumed initial infection with HIV. He suspended therapy on two occasions due to other concurrent infections, and while he had a rebound in viral load after the first treatment interruption, his viral load remained "undetectable" (less than 500 copies/ml) during the second interruption, which lasted 16 days. He stopped treatment completely after 176 days, and his viral load has remained undetectable for over 18 months.
HIV, however, has been found in his lymph nodes and at a low frequency in his circulating T-cells. This patient has had evidence of vigorous immune responses to HIV. Taken together, it is not clear whether the control of HIV replication in this patient is due to one or more of the following factors: his very early treatment after initial infection, the specific antiretroviral regimen used, and/or intermittent therapy. Perhaps intermittent exposure to HIV antigens during dose interruptions could act like a vaccine and stimulate the body's immune response to the virus, enabling the immune system to control viral replication.
At the Chicago conference, Dr. Lori's group reported preliminary findings which hinted that intermittent therapy may be worthy of further exploration. The researchers treated three patients with stable baseline viral loads ranging from about 23,000 to over 700,000 copies/ml with HAART regimens containing hydroxyurea.
Treatment was given for three weeks, followed by a one-week interruption, and then for two cycles of three months of treatment followed by interruption until viral load rebounded to over 5,000 at which time treatment was reintroduced. Viral load fell back to less than 400 each time treatment was re-started. Interestingly, the time to rebound of viral load to over 5,000 increased with each dose interruption. Rebound time lasted an average of seven days during the first interruption and increased to 14 days and 37 days for the second and third. Parallel studies in rhesus monkeys acutely infected with the related simian immunodeficiency virus (SIV) yielded similar results.
These very preliminary results in only a few patients must be interpreted with caution. In no way do they prove that intermittent therapy or drug holidays are beneficial. Randomized, controlled clinical trials to compare continuous with intermittent therapy are needed and are being planned.
Marshall Glesby is director of Chelsea's Cornell Clinical Trials Unit, and asst. professor of medicine at Cornell University Medical College.