Is autovaccination dead?

"Long-term non-progressors have a strong cellular anti-HIV immune response." Thus starts many a review article on HIV immunology. And many readers therefore conclude, "After induction of a strong anti-HIV immune response, an ordinary patient with HIV infection will be transformed into a long-term non-progressor."

But how to induce a strong anti-HIV immune response? In the absence of effective HIV vaccines, the virus itself will have to do. Stop HAART, have the virus rebound, start treatment again, and repeat the stop-start cycle a few times. This schedule would expose patients to several pulses of HIV antigen, and presumably jolt the immune system into responding. Such "autovaccination" would be followed by withdrawal of HAART. Hopefully, the newly stimulated immune system would then contain the virus without drugs.

Too good to be true? Despite widespread skepticism, data from experiments in monkeys with Simian Immunodeficiency Virus (SIV), as well as isolated case reports, suggested that the idea might have merit.

It has now been put to the test in 133 patients recruited in Switzerland and Spain. HAART was interrupted for 2 weeks, then restarted and given for 8 weeks. The off-on cycle was repeated 4 times. After week 40, treatment was suspended indefinitely, unless CD4 T cell counts dropped dangerously or viral loads became excessive. The level of post-trial viremia after 12 weeks off therapy was measured and compared to the level before HAART, with the hope that viremia post-trial would be lower than pre-HAART.

Anti-HIV immune responses were indeed stimulated, as measured by the number of lymphocytes reacting to HIV antigens, which rose from 300 to 2000 per million. But the treatment interruptions did not significantly influence the post-trial viremia, which rose close to the level before HAART. Moreover, a strong anti-HIV immune response in no way predicted low post-trial viremia; to the contrary, patients with strong stimulation of anti-HIV immune responses tended to have higher, not lower, viremia after interruption of therapy.

These results provide strong arguments against the hypothesis that "autovaccination" helps to improve immune control of HIV. They also cast a shadow of doubt on the prospects of therapeutic vaccination. Indeed, the immune response induced by the treatment interruptions was stronger than the most vigorous immune responses induced so far by experimental vaccines, and of course, the "vaccine" and the "challenge" viruses were exactly the same--no worries regarding mutants and variations in subtypes! If such an immune response is ineffective, so would a less vigorous response induced by a heterologous therapeutic vaccine.

Is autovaccination dead? Pretty much so, at least for the vast majority of patients who started therapy during chronic HIV infection. The concept of intermittent therapy may still have value--not because it provides effective autovaccination, but because lowering exposure to HAART might diminish side effects and diminish costs. Prospective trials comparing intermittent with continous treatment are ongoing.

Bernard Hirschel, MD

Division des Maladies Infectieuses, University Hospital, Geneva, Switzerland